Uploaded by Joseph Donaldson

fundamentals of applied electromagnetics, 8th edition

advertisement
We dedicate this book to
Jean and Ann Lucia.
Preface to Eighth Edition
Building on the core content and style of its predecessor, this
eighth edition (8/e) of Applied Electromagnetics includes features designed to help students develop deep understanding of
electromagnetic concepts and applications. Prominent among
them is a set of 52 web-based simulation modules∗ that allow
the user to interactively analyze and design transmission line
circuits; generate spatial patterns of the electric and magnetic
fields induced by charges and currents; visualize in 2-D and
3-D space how the gradient, divergence, and curl operate on
spatial functions; observe the temporal and spatial waveforms
of plane waves propagating in lossless and lossy media;
calculate and display field distributions inside a rectangular
waveguide; and generate radiation patterns for linear antennas and parabolic dishes. These are valuable learning tools;
we encourage students to use them and urge instructors to
incorporate them into their lecture materials and homework
assignments.
NEW TO THIS EDITION
• Additional exercises
• Updated Technology Briefs
• Enhanced figures and images
• New/updated end-of-chapter problems
* The interactive modules and Technology Briefs
can be found at the book companion website:
em8e.eecs.umich.edu.
ACKNOWLEDGMENTS
As authors, we were blessed to have worked on this book with
the best team of professionals: Richard Carnes, Leland Pierce,
Janice Richards, Rose Kernan, and Paul Mailhot. We are
exceedingly grateful for their superb support and unwavering
dedication to the project.
Additionally, by enhancing the book’s graphs and illustrations and by expanding the scope of topics of the Technology
Briefs, additional bridges between electromagnetic fundamentals and their countless engineering and scientific applications
are established.
We enjoyed working on this book. We hope you enjoy
learning from it.
FAWWAZ T. U LABY
U MBERTO R AVAIOLI
iv
v
CONTENT
This book begins by building a bridge between what should
be familiar to a third-year electrical engineering student and
the electromagnetics (EM) material covered in the book. Prior
to enrolling in an EM course, a typical student will have
taken one or more courses in circuits. He or she should be
familiar with circuit analysis, Ohm’s law, Kirchhoff’s current
and voltage laws, and related topics.
Transmission lines constitute a natural bridge between electric circuits and electromagnetics. Without having to deal with
vectors or fields, the student will use already familiar concepts
to learn about wave motion, the reflection and transmission
of power, phasors, impedance matching, and many of the
properties of wave propagation in a guided structure. All of
these newly learned concepts will prove invaluable later (in
Chapters 7 through 9) and will facilitate the learning of how
plane waves propagate in free space and in material media.
Transmission lines are covered in Chapter 2, which is preceded
in Chapter 1 with reviews of complex numbers and phasor
analysis.
The next part of this book, contained in Chapters 3
through 5, covers vector analysis, electrostatics, and magnetostatics. The electrostatics chapter begins with Maxwell’s equa-
tions for the time-varying case, which are then specialized to
electrostatics and magnetostatics. These chapters will provide
the student with an overall framework for what is to come
and show him or her why electrostatics and magnetostatics are
special cases of the more general time-varying case.
Chapter 6 deals with time-varying fields and sets the stage
for the material in Chapters 7 through 9. Chapter 7 covers
plane-wave propagation in dielectric and conducting media,
and Chapter 8 covers reflection and transmission at discontinuous boundaries and introduces the student to fiber optics,
waveguides, and resonators. In Chapter 9, the student is introduced to the principles of radiation by currents flowing in
wires, such as dipoles, as well as to radiation by apertures,
such as a horn antenna or an opening in an opaque screen
illuminated by a light source.
To give the student a taste of the wide-ranging applications
of electromagnetics in today’s technological society, Chapter 10 concludes this book with presentations of two system
examples: satellite communication systems and radar sensors.
The material in this book was written for a two-semester
sequence of six credits, but it is possible to trim it down to
generate a syllabus for a one-semester, four-credit course. The
accompanying table provides syllabi for each of these options.
Suggested Syllabi
1
2
3
4
5
6
7
8
9
10
Chapter
Introduction:
Waves and Phasors
Transmission Lines
Vector Analysis
Electrostatics
Magnetostatics
Exams
Maxwell’s Equations
for Time-Varying Fields
Plane-Wave Propagation
Wave Reflection
and Transmission
Radiation and Antennas
Satellite Communication
Systems and Radar Sensors
Exams
Two-Semester Syllabus
6 credits (42 contact hours per semester)
Sections
Hours
All
4
All
All
All
All
12
8
8
7
3
42
6
2-1 to 2-8 and 2-11
All
4-1 to 4-10
5-1 to 5-5 and 5-7 to 5-8
8
8
6
5
2
6-1 to 6-3, and 6-6
3
All
All
7
9
7-1 to 7-4, and 7-6
8-1 to 8-3, and 8-6
6
7
All
All
10
5
9-1 to 9-6
None
6
—
3
40
2
Total
1
56
0
Total for first semester
All
Total for second semester
Extra Hours
One-Semester Syllabus
4 credits (56 contact hours)
Sections
Hours
All
4
vi
MESSAGE TO THE STUDENT
ACKNOWLEDGMENTS
The web-based interactive modules of this book were developed with you, the student, in mind. Take the time to use them
in conjunction with the material in the textbook. The multiplewindow feature of electronic displays makes it possible to
design interactive modules with “help” buttons to guide you
through the solution of a problem when needed. Video animations can show you how fields and waves propagate in time and
space, how the beam of an antenna array can be made to scan
electronically, and how current is induced in a circuit under
the influence of a changing magnetic field. The modules are a
useful resource for self-study. You can find them at the book
companion website em8e.eecs.umich.edu. Use them!
Special thanks are due to our reviewers for their valuable
comments and suggestions. They include Constantine Balanis
of Arizona State University, Harold Mott of the University
of Alabama, David Pozar of the University of Massachusetts,
S. N. Prasad of Bradley University, Robert Bond of the New
Mexico Institute of Technology, Mark Robinson of the University of Colorado at Colorado Springs, and Raj Mittra of
the University of Illinois. I appreciate the dedicated efforts of
the staff at Prentice Hall, and I am grateful for their help in
shepherding this project through the publication process in a
very timely manner.
BOOK COMPANION WEBSITE
Throughout the book, we use the symbol EM to denote the book
companion website em8e.eecs.umich.edu, which contains a
wealth of information and tons of useful tools.
FAWWAZ T. U LABY
List of
Technology Briefs
TB1
TB2
TB3
TB4
TB5
TB6
TB7
TB8
TB9
LED Lighting
18
Solar Cells
35
Microwave Ovens
80
EM Cancer Zappers
117
Global Positioning System
151
X-Ray Computed Tomography 159
Resistive Sensors
186
Supercapacitors as Batteries
191
Capacitive Sensors
203
TB10
TB11
TB12
TB13
TB14
TB15
TB16
TB17
vii
Electromagnets
247
Inductive Sensors
262
EMF Sensors
290
RFID Systems
313
Liquid Crystal Display (LCD) 320
Lasers
347
Bar-Code Readers
358
Health Risks of EM Fields
405
Contents
Contents
Preface
iv
List of Technology Briefs
vii
List of Modules
xii
Photo Credits
xiii
1 Introduction: Waves and Phasors
1-1 Historical Timeline . . . . . . .
1-2 Dimensions, Units, and Notation
1-3 The Nature of Electromagnetism
TB1 LED Lighting . . . . . . . . . .
1-4 Traveling Waves . . . . . . . . .
1-5 The Electromagnetic Spectrum .
1-6 Review of Complex Numbers . .
TB2 Solar Cells . . . . . . . . . . . .
1-7 Review of Phasors . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2 Transmission Lines
2-1 General Considerations . . . . . . . . . . . . . . . . . .
2-2 Lumped-Element Model . . . . . . . . . . . . . . . . . .
2-3 Transmission-Line Equations . . . . . . . . . . . . . . .
2-4 Wave Propagation on a Transmission Line . . . . . . . .
2-5 The Lossless Microstrip Line . . . . . . . . . . . . . . .
2-6 The Lossless Transmission Line: General Considerations
2-7 Wave Impedance of the Lossless Line . . . . . . . . . . .
2-8 Special Cases of the Lossless Line . . . . . . . . . . . .
TB3 Microwave Ovens . . . . . . . . . . . . . . . . . . . . .
2-9 Power Flow on a Lossless Transmission Line . . . . . . .
2-10 The Smith Chart . . . . . . . . . . . . . . . . . . . . . .
2-11 Impedance Matching . . . . . . . . . . . . . . . . . . .
2-12 Transients on Transmission Lines . . . . . . . . . . . . .
viii
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
3
11
11
18
22
30
31
35
38
.
.
.
.
.
.
.
.
.
.
.
.
.
46
47
50
53
54
59
62
71
74
80
82
84
94
108
CONTENTS
ix
TB4 EM Cancer Zappers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
3 Vector Analysis
3-1 Basic Laws of Vector Algebra . . . . . . . . .
3-2 Orthogonal Coordinate Systems . . . . . . . .
3-3 Transformations between Coordinate Systems
3-4 Gradient of a Scalar Field . . . . . . . . . . .
TB5 Global Positioning System . . . . . . . . . .
3-5 Divergence of a Vector Field . . . . . . . . .
3-6 Curl of a Vector Field . . . . . . . . . . . . .
TB6 X-Ray Computed Tomography . . . . . . . .
3-7 Laplacian Operator . . . . . . . . . . . . . .
4 Electrostatics
4-1 Maxwell’s Equations . . . . . .
4-2 Charge and Current Distributions
4-3 Coulomb’s Law . . . . . . . . .
4-4 Gauss’s Law . . . . . . . . . . .
4-5 Electric Scalar Potential . . . . .
TB7 Resistive Sensors . . . . . . . .
4-6 Conductors . . . . . . . . . . . .
TB8 Supercapacitors as Batteries . . .
4-7 Dielectrics . . . . . . . . . . . .
4-8 Electric Boundary Conditions . .
TB9 Capacitive Sensors . . . . . . . .
4-9 Capacitance . . . . . . . . . . .
4-10 Electrostatic Potential Energy . .
4-11 Image Method . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
130
131
137
143
147
151
153
157
159
162
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
172
173
173
176
180
183
186
190
191
197
200
203
209
213
215
5 Magnetostatics
5-1 Magnetic Forces and Torques . . .
5-2 The Biot–Savart Law . . . . . . .
5-3 Maxwell’s Magnetostatic Equations
5-4 Vector Magnetic Potential . . . . .
TB10 Electromagnets . . . . . . . . . .
5-5 Magnetic Properties of Materials .
5-6 Magnetic Boundary Conditions . .
5-7 Inductance . . . . . . . . . . . . .
5-8 Magnetic Energy . . . . . . . . . .
TB11 Inductive Sensors . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
227
228
236
242
246
247
251
254
256
261
262
6 Maxwell’s Equations for Time-Varying Fields
272
6-1 Faraday’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
6-2 Stationary Loop in a Time-Varying Magnetic Field . . . . . . . . . . . . . . . . . . . . . . 274
6-3 The Ideal Transformer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
x
CONTENTS
6-4
6-5
6-6
6-7
6-8
6-9
6-10
TB12
6-11
Moving Conductor in a Static Magnetic Field . . . .
The Electromagnetic Generator . . . . . . . . . . . .
Moving Conductor in a Time-Varying Magnetic Field
Displacement Current . . . . . . . . . . . . . . . . .
Boundary Conditions for Electromagnetics . . . . . .
Charge–Current Continuity Relation . . . . . . . . .
Free-Charge Dissipation in a Conductor . . . . . . .
EMF Sensors . . . . . . . . . . . . . . . . . . . . . .
Electromagnetic Potentials . . . . . . . . . . . . . .
7 Plane-Wave Propagation
7-1 Time-Harmonic Fields . . . . . . . . . . .
7-2 Plane-Wave Propagation in Lossless Media
7-3 Wave Polarization . . . . . . . . . . . . .
TB13 RFID Systems . . . . . . . . . . . . . . .
7-4 Plane-Wave Propagation in Lossy Media .
TB14 Liquid Crystal Display (LCD) . . . . . . .
7-5 Current Flow in a Good Conductor . . . .
7-6 Electromagnetic Power Density . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
279
283
284
285
287
288
289
290
292
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
301
303
304
309
313
317
320
325
328
8 Wave Reflection and Transmission
8-1 Wave Reflection and Transmission at Normal Incidence
TB15 Lasers . . . . . . . . . . . . . . . . . . . . . . . . . .
8-2 Snell’s Laws . . . . . . . . . . . . . . . . . . . . . . .
8-3 Fiber Optics . . . . . . . . . . . . . . . . . . . . . . .
8-4 Wave Reflection and Transmission at Oblique Incidence
TB16 Bar-Code Readers . . . . . . . . . . . . . . . . . . . .
8-5 Reflectivity and Transmissivity . . . . . . . . . . . . .
8-6 Waveguides . . . . . . . . . . . . . . . . . . . . . . .
8-7 General Relations for E and H . . . . . . . . . . . . .
8-8 TM Modes in Rectangular Waveguide . . . . . . . . .
8-9 TE Modes in Rectangular Waveguide . . . . . . . . . .
8-10 Propagation Velocities . . . . . . . . . . . . . . . . . .
8-11 Cavity Resonators . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
337
338
347
349
351
353
358
361
364
366
367
370
371
374
.
.
.
.
.
.
.
.
384
387
390
397
400
401
403
405
408
9 Radiation and Antennas
9-1 The Hertzian Dipole . . . . . . . . . .
9-2 Antenna Radiation Characteristics . .
9-3 Half-Wave Dipole Antenna . . . . . .
9-4 Dipole of Arbitrary Length . . . . . .
9-5 Effective Area of a Receiving Antenna
9-6 Friis Transmission Formula . . . . . .
TB17 Health Risks of EM Fields . . . . . . .
9-7 Radiation by Large-Aperture Antennas
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
CONTENTS
9-8
9-9
9-10
9-11
xi
Rectangular Aperture with Uniform Aperture Distribution
Antenna Arrays . . . . . . . . . . . . . . . . . . . . . .
N-Element Array with Uniform Phase Distribution . . . .
Electronic Scanning of Arrays . . . . . . . . . . . . . . .
10 Satellite Communication Systems and Radar Sensors
10-1 Satellite Communication Systems . . . . . . . . .
10-2 Satellite Transponders . . . . . . . . . . . . . . .
10-3 Communication-Link Power Budget . . . . . . .
10-4 Antenna Beams . . . . . . . . . . . . . . . . . .
10-5 Radar Sensors . . . . . . . . . . . . . . . . . . .
10-6 Target Detection . . . . . . . . . . . . . . . . . .
10-7 Doppler Radar . . . . . . . . . . . . . . . . . . .
10-8 Monopulse Radar . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
410
412
419
421
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
434
435
436
439
440
441
443
445
447
Appendix A: Symbols, Quantities, and Units
451
Appendix B: Material Constants of Some Common Materials
454
Appendix C: Mathematical Formulas
457
Appendix D: Fundamental Constants and Units
460
Appendix E: Answers to Selected Problems
462
Bibliography
466
Index
468
List of Modules
To access and exercise the following electronic modules, go to em8e.eecs.umich.edu.
1.1
1.2
1.3
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
3.1
3.2
3.3
3.4
4.1
4.2
4.3
4.4
5.1
5.2
5.3
5.4
Sinusoidal Waveforms
Traveling Waves
Phase Lead/Lag
Two-Wire Line
Coaxial Line
Lossless Microstrip Line
Transmission Line Simulator
Wave and Input Impedance
Interactive Smith Chart
Quarter-Wavelength Transformer
Discrete Element Matching
Single-Stub Tuning
Transient Response
Points and Vectors
Gradient
Divergence
Curl
Fields Due to Charges
Charges in Adjacent Dielectrics
Charges above a Conducting Plane
Charges near a Conducting Sphere
Electron Motion in Static Fields
Magnetic Fields Due to Line Sources
Magnetic Field of a Current Loop
Magnetic Forces between Two
Parallel Conductors
27
29
33
56
57
61
70
74
96
106
107
108
116
141
150
156
162
195
208
210
211
231
235
239
241
6.1
6.2
6.3
7.1
7.2
7.3
7.4
7.5
7.6
8.1
8.2
8.3
8.4
8.5
9.1
9.2
9.3
9.4
9.5
9.6
9.7
9.8
xii
Circular Loop in Time-Varying
Magnetic Field
Rotating Wire Loop in Constant
Magnetic Field
Displacement Current
Linking E to H
Plane Wave
Polarization I
Polarization II
Wave Attenuation
Current in a Conductor
Incidence on Perfect Conductor
Multimode Step-Index Optical Fiber
Oblique Incidence
Oblique Incidence in Lossy Medium
Rectangular Waveguide
Hertzian Dipole
Linear Dipole Antenna
Detailed Analysis of Linear Antenna
Large Parabolic Reflector
Two-Dipole Array
Detailed Analysis of Two-Dipole Array
N-Element Array
Uniform Dipole Array
277
282
287
308
310
317
318
324
327
346
353
363
364
376
391
401
402
413
417
418
422
425
Photo Credits
p. 4 (Ch 01-01A): Thales of Miletus (624–546 BC), Photo
Researchers, Inc./Science Source
p. 6 (Ch 01-02E): De Forest seated at his invention, the
radio-telephone, called the Audion, New York Public
Library/Science Source
p. 4 (Ch 01-01B): Isaac Newton, World History Archive/Alamy
Stock Photo
p. 6 (Ch 01-02F): The staff of KDKA broadcast reports of the
1920 presidential election, Bettmann/Getty Images
p. 4 (Ch 01-01C): Benjamin West, Benjamin Franklin Drawing
Electricity from the Sky, Painting/Alamy Stock Photo
p. 7 (Ch 01-02G): This bottle-like object is a Cathode Ray tube
which forms the receiver of the new style television invented
by Dr. Vladimir Zworykin, Westinghouse research engineer,
who is holding it, Album/Alamy Stock Photo
p. 4 (Ch 01-01D): Replica of the Voltaic pile invented by
Alessandro Volta 1800, Gio.tto/Shutterstock
p. 4 (Ch 01-01E): Hans Christian Ørsted, Danish Physicist, New
York Public Library/Science Source
p. 7 (Ch 01-02H): Radar in operation in the Second World War,
Library of Congress Department of Prints and Photographs
[LC-USZ62-101012]
p. 4 (Ch 01-01F): Andre-Marie Ampére, Nickolae/Fotolia
p. 5 (Ch 01-01G): Michael Faraday, Nicku/Shutterstock
p. 7 (Ch 01-02I): Shockley, Brattain, and Bardeen with an
apparatus used in the early investigations which led to the
invention of the transistor, New York Public Library/Science
Source
p. 5 (Ch 01-01H): James Clerk Maxwell (1831–1879),
Nicku/Shutterstock
p. 5 (Ch 01-01I): Heinrich Rudolf Hertz, New York Public
Library/Science Source
p. 7 (Ch 01-02J): A Photograph of Jack Kilby’s Model of the First
Working Integrated Circuit Ever Built circa 1958,
Fotosearch/Archive Photos/Getty Images
p. 5 (Ch 01-01J): Nicola Tesla, NASA
p. 5 (Ch 01-01K): Early X-Ray of Hand, Science History
Images/Alamy Stock Photo
p. 7 (Ch 01-02K): Shown here is the 135-foot rigidized inflatable
balloon satellite undergoing tensile stress test in a dirigible
hanger at Weekesville, North Carolina, NASA
p. 5 (Ch 01-01M): Albert Einstein, LOC/Science Source
p. 6 (Ch 01-02A): Telegraph, Morse apparatus, vintage engraved
illustration, Morphart Creation/Shutterstock
p. 7 (Ch 01-02L): Pathfinder on Mars, JPL/NASA
p. 8 (Ch 01-03A): Abacus isolated on white, Sikarin
Supphatada/Shutterstock
p. 6 (Ch 01-02B): Thomas Alva Edison with his ‘Edison Effect’
Lamps, Education Images/Universal Images Group/Getty
Images, Inc.
p. 8 (Ch 01-03B): Pascaline; a mechanical calculator invented by
Blaise Pascal in 1642, New York Public Library/Science
Source
p. 6 (Ch 01-02C): Replica of an early type of telephone made by
Scottish-born telephony pioneer Alexander Graham Bell
(1847–1922), Science & Society Picture Library/Getty
Images
p. 8 (Ch 01-03C): Original Caption: Portrait of American
electrical engineer Vannevar Bush, Bettmann/Getty Images
p. 6 (Ch 01-02D): Guglielmo Marconi, Pach Brothers/Library of
Congress Prints and Photographs Division
[LC-USZ62-39702]
p. 8 (Ch 01-03D): J. Presper Eckert and John W. Mauchly, are
pictured with the Electronic Numerical Integrator and
Computer (ENIAC) in this undated photo from the University
xiii
xiv
of Pennsylvania Archives, University of Pennsylvania/AP
images
p. 8 (Ch 01-03E): Description: DEC PDP-1 computer, on display
at the Computer History Museum, USA, Volker
Steger/Science Source
p. 9 (Ch 01-03F): Classic Antique Red LED Diode Calculator,
James Brey/E+/Getty Images
p. 9 (Ch 01-03G): Apple I computer. This was released in April
1976 at the Homebrew Computer Club, USA, Mark
Richards/ZUMAPROSS.com/Alamy Stock Photo
p. 9 (Ch 01-03H): UNITED STATES—DECEMBER 07: The
IBM Personal Computer System was introduced to the
market in early 1981, Science & Society Picture
Library/Getty Images, Inc.
p. 9 (Ch 01-03I): NEW YORK, UNITED STATES: Chess
enthusiasts watch world chess champion Garry Kasparov on a
television monitor as he holds his head in his hands, Adam
Nadel/AP/Shutterstock
p. 10 (Fig. 1-2(a)): The Very Large Array of Radio Telescopes,
VLA, NRAO/NASA
p. 10 (Fig. 1-2(b)): SCaN’s Benefits to Society—Global
Positioning System, Jet Propulsion Laboratory/NASA
p. 117 (TF4-2): Setup for a percutaneous microwave ablation
procedure shows three single microwave applicators
connected to three microwave generators, Radiological
Society of North America (RSNA)
p. 118 (TF4-3): Bryan Christie Design and the Institute of Electrical
and Electronics Engineers. IEEE Spectrum by Institute of
Electrical and Electronics Engineers. Reproduced with
permission of Institute of Electrical and Electronics
Engineers, in the format Republish in a book via Copyright
Clearance Center
p. 150 (Mod. 3-2): Screenshot: Gradient. Graphics created with
R
Wolfram Mathematica°
used with permission
p. 151 (TF5-1): Touchscreen smartphone with GPS navigation
isolated on white reflective background, Oleksiy
Mark/Shutterstock
p. 151 (TF5-2): SCaN’s Benefits to Society—Global Positioning
System, U.S. Department of Defense as a navigation tool for
military, developed in 1980
p. 152 (TF5-3): SUV, Konstantin/Fotolia
p. 156 (Mod. 3-3): Screenshot: Divergence, Graphics created with
R
Wolfram Mathematica°
used with permission
p. 10 (Fig. 1-2(c)): Motor, ABB
p. 159 (TF6-1): X-ray of pelvis and spinal column, Cozyta
Moment/Getty Images, Inc.
p. 10 (Fig. 1-2(d) and page 338 (Fig. TF14-04)): TV on white
background, Fad82/Shutterstock
p. 159 (TF6-2): CT scan advance technology for medical diagnosis,
Tawesit/Fotolia
p. 10 (Fig. 1-2(e)): Nuclear Propulsion Through Direct Conversion
of Fusion Energy, John Slough/NASA
p. 160 (TF6-3(c)): Digitally enhanced CT scan of a normal brain in
transaxial (horizontal) section, Scott Camazine/Science
Source
p. 10 (Fig. 1-2(f)): Tracking station has bird’s eye view on VAFB,
Ashley Tyler/US Air Force
p. 10 (Fig. 1-2(g)): Glass Fiber Cables, Valentyn Volkov/123RF
p. 10 (Fig. 1-2(h)): Touchscreen smartphone, Oleksiy
Mark/Shutterstock
p. 10 (Fig. 1-2(i)): Line Art: Electromagnetics is at the heart of
numerous systems and applications. Source: Based on IEEE
Spectrum
p. 162 (Mod. 3-4): Screenshot: Curl, Graphics created with Wolfram
R
Mathematica°
used with permission
p. 186 (TF7-1): Most cars use over 100 sensors, National Motor
Museum/Shutterstock
p. 191 (TF8-1): Examples of supercapacitors. Source: Vladimir
Zhupanenko/Shutterstock
p. 191 (TF8-2(a)): High-speed train in motion, Metlion/Fotolia
p. 18 (TF1-1(a)): Lightbulb, Chones/Fotolia
p. 191 (TF8-2(b)): Cordless Drill, Derek Hatfield/Shutterstock
p. 18 (TF1-1(b)): Fluorescent bulb, Wolf1984/Fotolia
p. 191 (TF8-2(c)): The 2006 BMW X3 Concept Gasoline Electric
Hybrid uses high-performance capacitors (or “Super Caps”)
to store and supply electric energy to the vehicle’s Active
Transmission, Passage/Car Culture/Getty Images
p. 18 (TF1-1(c)): 3d render of an unbranded screw-in LED lamp,
isolated on a white background, Marcello Bortolino/Getty
Images, Inc.
p. 19 (TF1-3): Line Art: Lighting efficiency. Source: Based on
National Research Council, 2009
p. 32 (Fig. 1-17): Individual bands of the radio spectrum and their
primary allocations in the U.S. [See expandable version on
book website: em8e.eecs.umich.edu.] Source: U.S.
Department of Commerce
p. 59 (Fig. 2-10(c)): Circuit board, Gabriel Rebeiz
p. 117 (TF4-1): Microwave ablation for cancer liver treatment,
Radiological Society of North America (RSNA)
p. 191 (TF8-2(d)): LED Electric torch—laser pointer isolated on
white background, Artur Synenko/Shutterstock
p. 206 (TF9-6): Elements of a fingerprint matching system. Bryan
Christie Design and the Institute of Electrical and Electronics
Engineers. IEEE Spectrum by Institute of Electrical and
Electronics Engineers. Reproduced with permission of
Institute of Electrical and Electronics Engineers, in the format
Republish in a book via Copyright Clearance Center
p. 206 (TF9-7): Line Art: Fingerprint representation, Source:
Courtesy of Dr. M. Tartagni, University of Bologna, Italy
xv
p. 249 (TF10-5(a)): CHINA—JUNE 20: A maglev train awaits
departure in Shanghai, China, on Saturday, June 20, 2009,
Qilai Shen/Bloomberg/Getty Images
p. 249 (TF10-5(b) and (c)): Line Art: Magnetic trains—(b) internal
workings of the Maglev train. Used with permission—Amy
Mast, “Maglev trains are making history right now.” Flux,
Volume 3, Issue 1, National High Magnetic Field Laboratory
p. 291 (TF12-2): Ultrasonic Transducer, NDT Resource
p. 313 (TF13-1): Jersey cow on pasture, Lakeview
Images/Shutterstock
p. 314 (TF13-2): Line Art: How an RFID system works is illustrated
through this EZ-Pass example: Tag. Source: Cary
Wolinsky/Cavan Images/Alamy Stock Photo
p. 347 (TF15-1(a)): Optical Computer Mouse, Satawat
Anukul/123RF
p. 347 (TF15-1(b)): Laser eye surgery, Will & Deni
McIntyre/Science Source
p. 347 (TF15-1(c)): Laser Star Guide, NASA
p. 405 (TF17-1(a)): Smiling woman using computer,
Edbockstock/Fotolia
p. 405 (TF17-1(b)): Vector silhouette of Power lines and electric
pylons, Ints Vikmanis/Alamy Stock Photo
p. 405 (TF17-1(c)): Telecommunications tower, Poliki/Fotolia
p. 414 (Fig. 9-25): The AN/FPS-85 Phased Array Radar Facility in
the Florida panhandle, near the city of Freeport, NASA
Chapter
1
Introduction:
Waves and Phasors
Chapter Contents
1-1
1-2
1-3
TB1
1-4
1-5
1-6
TB2
1-7
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Overview, 2
Historical Timeline, 3
Dimensions, Units, and Notation, 11
The Nature of Electromagnetism, 11
LED Lighting, 18
Traveling Waves, 22
The Electromagnetic Spectrum, 30
Review of Complex Numbers, 31
Solar Cells, 35
Review of Phasors, 38
Chapter 1 Summary, 42
Problems, 42
1. Describe the basic properties of electric and magnetic
forces.
2. Ascribe mathematical formulations to sinusoidal waves
traveling in both lossless and lossy media.
3. Apply complex algebra in rectangular and polar forms.
4. Apply the phasor-domain technique to analyze circuits
driven by sinusoidal sources.
1
2
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
2-D pixel array
Liquid crystal
678
Unpolarized light
Exit polarizer
Entrance polarizer
Molecular spiral
LCD display
Figure 1-1 2-D LCD array.
Overview
Liquid crystal displays (LCDs) have become integral parts
of many electronic consumer products, ranging from alarm
clocks and cell phones to laptop computers and television systems. LCD technology relies on special electrical and optical
properties of a class of materials known as liquid crystals,
which are neither pure solids nor pure liquids but rather a
hybrid of both. The molecular structure of these materials is
such that when light travels through them, the polarization
of the emerging light depends on whether or not a voltage
exists across the material. Consequently, when no voltage is
applied, the exit surface appears bright. Conversely, when a
voltage of a certain level is applied across the LCD material,
no light passes through it, resulting in a dark pixel. In-between
voltages translate into a range of grey levels. By controlling the
voltages across individual pixels in a two-dimensional array, a
complete image can be displayed (Fig. 1-1). Color displays are
composed of three subpixels with red, green, and blue filters.
◮ The polarization behavior of light in a LCD is a
prime example of how electromagnetics is at the heart of
electrical and computer engineering. ◭
The subject of this book is applied electromagnetics (EM),
which encompasses the study of both static and dynamic
electric and magnetic phenomena and their engineering
applications. Primary emphasis is placed on the fundamental
properties of dynamic (time-varying) electromagnetic fields
because of their greater relevance to practical applications,
including wireless and optical communications, radar, bioelectromagnetics, and high-speed microelectronics. We study
wave propagation in guided media, such as coaxial transmission lines, optical fibers, and waveguides; wave reflection and
transmission at interfaces between dissimilar media; radiation
by antennas, and several other related topics. The concluding
chapter is intended to illustrate a few aspects of applied EM
through an examination of design considerations associated
with the use and operation of radar sensors and satellite
communication systems.
We begin this chapter with a chronology of the history of
electricity and magnetism. Next, we introduce the fundamental
electric and magnetic field quantities of electromagnetics, as
well as their relationships to each other and to the electric
charges and currents that generate them. These relationships
constitute the underpinnings of the study of electromagnetic
phenomena. Then, in preparation for the material presented in
Chapter 2, we provide short reviews of three topics: traveling
waves, complex numbers, and phasors, which are all useful in
solving time-harmonic problems.
1-1 HISTORICAL TIMELINE
1-1 Historical Timeline
The history of EM may be divided into two overlapping eras.
The first is the classical era, during which the fundamental
laws of electricity and magnetism were discovered and formulated. Building on these formulations, the modern era of the
past 100 years ushered in the birth of the field of applied EM
as we know it today.
1-1.1 EM in the Classical Era
Chronology 1-1 provides a timeline for the development of
electromagnetic theory in the classical era. It highlights those
discoveries and inventions that have impacted the historical
development of EM in a very significant way, even though the
selected discoveries represent only a small fraction of those
responsible for our current understanding of electromagnetics. As we proceed through this book, some of the names
highlighted in Chronology 1-1, such as those of Coulomb
and Faraday, will appear again as we discuss the laws and
formulations named after them.
The attractive force of magnetite was reported by the Greeks
some 2800 years ago. It was also a Greek, Thales of Miletus,
who first wrote about what we now call static electricity: He
described how rubbing amber caused it to develop a force
that could pick up light objects such as feathers. The term
“electric” first appeared in print around 1600 in a treatise
on the (electric) force generated by friction, authored by the
physician to Queen Elizabeth I, William Gilbert.
About a century later, in 1733, Charles-François du Fay
introduced the notion that electricity involves two types of
“fluids,” one “positive” and the other “negative,” and that
like-fluids repel and opposite-fluids attract. His notion of a
fluid is what we today call electric charge. The invention
of the capacitor in 1745, originally called the Leyden jar,
made it possible to store significant amounts of electric charge
in a single device. A few years later, in 1752, Benjamin
Franklin demonstrated that lightning is a form of electricity.
He transferred electric charge from a cloud to a Leyden jar via
a silk kite flown in a thunderstorm. The collective eighteenthcentury knowledge about electricity was integrated in 1785 by
Charles-Augustin de Coulomb, in the form of a mathematical
formulation characterizing the electrical force between two
charges in terms of their strengths and polarities and the
distance between them.
The year 1800 is noted for the development of the first electric battery by Alessandro Volta, and 1820 was a banner year
for discoveries about how electric currents induce magnetism.
This knowledge was put to good use by Joseph Henry, who
developed one of the earliest electromagnets and dc (direct
current) electric motors. Shortly thereafter, Michael Faraday
built the first electric generator (the converse of the electric
3
motor). Faraday, in essence, demonstrated that a changing
magnetic field induces an electric field (and hence a voltage).
The converse relation, namely that a changing electric field
induces a magnetic field, was first proposed by James Clerk
Maxwell in 1864 and then incorporated into his four (now)
famous equations in 1873.
◮ Maxwell’s equations represent the foundation of classical electromagnetic theory. ◭
Maxwell’s theory, which predicted the existence of electromagnetic waves, was not fully accepted by the scientific
community at that time. It was later verified experimentally
by means of radio waves by Heinrich Hertz in the 1880s. Xrays, another member of the EM family, were discovered in
1895 by Wilhelm Röntgen. In the same decade, Nikola Tesla
was the first to develop the ac motor, which was considered a
major advance over its predecessor, the dc motor.
Despite the advances made in the 19th century in our
understanding of electricity and magnetism and how to put
them to practical use, it was not until 1897 that the fundamental
carrier of electric charge, the electron, was identified and its
properties quantified by Joseph Thomson. The ability to eject
electrons from a material by shining electromagnetic energy,
such as light, on it is known as the photoelectric effect.
◮ To explain the photoelectric effect, Albert Einstein
adopted the quantum concept of energy that had been
advanced a few years earlier (1900) by Max Planck.
Symbolically, this step represents the bridge between the
classical and modern eras of electromagnetics. ◭
1-1.2 EM in the Modern Era
Electromagnetics play a role in the design and operation
of every conceivable electronic device, including the diode,
transistor, integrated circuit, laser, display screen, bar-code
reader, cell phone, and microwave oven, to name but a few.
Given the breadth and diversity of these applications (Fig. 1-2),
it is far more difficult to construct a meaningful timeline
for the modern era than for the classical era. That said,
one can develop timelines for specific technologies and link
their milestone innovations to EM. Chronologies 1-2 and 1-3
present timelines for the development of telecommunications
and computers, technologies that have become integral parts
of today’s societal infrastructure. Some of the entries in these
chronologies refer to specific inventions, such as the telegraph,
the transistor, and the laser. The operational principles and
capabilities of some of these technologies are highlighted in
4
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Chronology 1-1: TIMELINE FOR ELECTROMAGNETICS IN THE CLASSICAL ERA
Electromagnetics in the Classical Era
ca. 900 Legend has it that, while walking across a field in northern
Greece, a shepherd named Magnus experiences a pull
BC
1752
Benjamin Franklin
(American) invents
the lightning rod and
demonstrates that
lightning is electricity.
1785
Charles-Augustin
de Coulomb (French)
demonstrates that the
electrical force between
charges is proportional to
the inverse of the square
of the distance between
them.
1800
Alessandro Volta (Italian)
develops the first electric
battery.
1820
Hans Christian Oersted
(Danish) demonstrates the
interconnection between
electricity and magnetism
through his discovery that
an electric current in a
wire causes a compass
needle to orient itself
perpendicular to
the wire.
1820
André-Marie Ampère (French)
notes that parallel currents in
wires attract each other and
opposite currents repel.
1820
Jean-Baptiste Biot (French)
and Félix Savart (French)
develop the Biot–Savart law
relating the magnetic field
induced by a wire segment
to the current flowing through it.
on the iron nails in his sandals by the black rock he is
standing on. The region was later named Magnesia and
the rock became known as magnetite [a form of iron with
permanent magnetism].
ca. 600 Greek philosopher Thales
describes how amber, after being
BC
rubbed with cat fur, can pick up
feathers [static electricity].
ca.
1000
1600
1671
1733
1745
Magnetic compass used as a
navigational device.
William Gilbert (English) coins the term electric after the
Greek word for amber (elektron), and observes that a
compass needle points north–south because the Earth
acts as a bar magnet.
Isaac Newton (English) demonstrates that white light is a
mixture of all the colors.
Charles-François du Fay (French) discovers that
electric charges are of two forms and that like charges
repel and unlike charges attract.
Pieter van Musschenbroek (Dutch) invents the Leyden
jar, which was the first electrical capacitor.
1-1 HISTORICAL TIMELINE
5
Chronology 1-1: TIMELINE FOR ELECTROMAGNETICS IN THE CLASSICAL ERA (continued)
Electromagnetics in the Classical Era
1827
Georg Simon Ohm (German) formulates Ohm's law
relating electric potential to current and resistance.
1827
Joseph Henry (American) introduces the concept of
inductance and builds one of the earliest electric motors.
He also assisted Samual Morse in the development
of the telegraph.
1831
Michael Faraday (English)
discovers that a changing
magnetic flux can induce
an electromotive force.
1835
Carl Friedrich Gauss (German) formulates Gauss's law
relating the electric flux flowing through an enclosed
surface to the enclosed electric charge.
1873
James Clerk Maxwell
(Scottish) publishes his
Treatise on Electricity and
Magnetism, in which he unites
the discoveries of Coulomb,
Oersted, Ampère, Faraday,
and others into four elegantly
constructed mathematical
equations, now known as
Maxwell’s Equations.
1887
Heinrich Hertz
(German) builds
a system that
can generate
electromagnetic
waves (at radio
frequencies) and
detect them.
1888
Nikola Tesla (American)
invents the ac (alternating
current) electric motor.
1895
Wilhelm Röntgen (German)
discovers X-rays. One of
his first X-ray images was
of the bones in his wife's
hands. [1901 Nobel prize
in physics.]
1897
Joseph John Thomson (English) discovers the electron
and measures its charge-to-mass ratio. [1906 Nobel prize
in physics.]
1905
Albert Einstein (German-American) explains the
photoelectric effect discovered earlier by Hertz in 1887.
[1921 Nobel prize in physics.]
6
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Chronology 1-2: TIMELINE FOR TELECOMMUNICATIONS
Telecommunications
1825
William Sturgeon
(English) develops
the multiturn
electromagnet.
1837
Samuel Morse
(American) patents the
electromagnetic telegraph
using a code of dots and
dashes to represent letters
and numbers.
1872
Thomas Edison (American)
patents the electric
typewriter.
1876
Alexander Graham Bell
(Scottish-American) invents
the telephone. The rotary dial
becomes available in 1890,
and by 1900, telephone
systems are installed in
many communities.
1887
Heinrich Hertz (German)
generates radio waves and
demonstrates that they
share the same properties
as light.
1887
Emil Berliner (American) invents the flat gramophone
disc, or record.
1896
Guglielmo Marconi (Italian)
files his first of many patents
on wireless transmission
by radio. In 1901, he
demonstrates radio telegraphy
across the Atlantic Ocean.
[1909 Nobel prize in physics,
shared with Karl Braun
(German).]
1897
Karl Braun (German) invents the cathode ray tube (CRT).
[1909 Nobel prize in physics, shared with Marconi.]
1902
Reginald Fessenden (American) invents amplitude
modulation for telephone transmission. In 1906, he
introduces AM radio broadcasting of speech and music
on Christmas Eve.
1912
Lee De Forest
(American)
develops the triode
tube amplifier for
wireless telegraphy.
Also in 1912, the
wireless distress
call issued by the
Titanic was heard
58 miles away by
the ocean liner
Carpathia, which
managed to rescue
705 Titanic passengers
3.5 hours later.
1919
Edwin Armstong (American) invents the
superheterodyne radio receiver.
1920
Birth of commercial radio
broadcasting; Westinghouse Corporation
establishes radio station
KDKA in Pittsburgh,
Pennsylvania.
1-1 HISTORICAL TIMELINE
7
Chronology 1-2: TIMELINE FOR TELECOMMUNICATIONS (continued)
Telecommunications
1923
Vladimir Zworykin
(Russian-American)
invents television. In
1926, John Baird (Scottish)
transmits TV images
over telephone wires
from London to Glasgow.
Regular TV broadcasting
began in Germany (1935),
England (1936), and the
United States (1939).
1958
Jack Kilby (American) builds first integrated circuit (IC) on
germanium and, independently, Robert Noyce (American)
builds first IC on silicon.
1960
Echo, the first passive
communication satellite, is
launched and successfully
reflects radio signals back
to Earth. In 1963, the first
communication satellite is
placed in geosynchronous orbit.
1926
Transatlantic telephone service between London and
New York.
1932
First microwave telephone link, installed (by Marconi)
between Vatican City and the Pope’s summer residence.
1933
Edwin Armstrong (American) invents frequency
modulation (FM) for radio transmission.
1969
ARPANET is established by the U.S. Department of
Defense and will evolve later into the Internet.
1935
Robert Watson-Watt
(Scottish) invents radar.
1979
1938
H. A. Reeves (American)
invents pulse code
modulation (PCM).
Japan builds the first cellular telephone network:
• 1983: Cellular phone networks start in the United States.
• 1990: Electronic beepers become common.
• 1995: Cell phones become widely available.
• 2002: Cell phone supports video and Internet.
1984
Worldwide Internet becomes operational.
1988
First transatlantic optical
fiber cable deployed between
the U.S. and Europe.
1997
The Mars Pathfinder sends
images to Earth.
Pager is introduced as a radio communication product in
hospitals and factories.
2004
Wireless communication is supported by many airports,
university campuses, and other facilities.
Narinder Kapany (Indian-American) demonstrates the
optical fiber as a low-loss, light-transmission medium.
2012
Smartphones worldwide exceed 1 billion.
1947
1955
1955
William Shockley,
Walter Brattain, and
John Bardeen (all
Americans) invent the
junction transistor at Bell
Labs [1956 Nobel prize
in physics].
8
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Chronology 1-3: TIMELINE FOR COMPUTER TECHNOLOGY
Computer Technology
ca 1100 The abacus is the earliest known calculating device.
BC
1614
John Napier (Scottish) develops the logarithm system.
1642
Blaise Pascal
(French) builds
the first adding
machine using
multiple dials.
1941
Konrad Zuze (German) develops the first programmable
digital computer, making use of binary arithmetic and
electric relays.
1945
John Mauchly and J. Presper Eckert (both American)
develop the ENIAC, which is the first all-electronic
computer.
1950
Yoshiro Nakama (Japanese) patents the floppy disk as a
magnetic medium for storing data.
1956
John Backus (American)
develops FORTRAN, which
is the first major programming
language.
1671
Gottfried von Leibniz (German) builds calculator that can
do both addition and multiplication.
1820
Charles Xavier Thomas de Colmar (French) builds the
Arithmometer: the first mass-produced calculator.
1958
Bell Labs develops the modem.
1885
Dorr Felt (American) invents and markets a key-operated
adding machine (and adds a printer in 1889).
1960
1930
Vannevar Bush (American) develops the differential analyzer,
which is an analog computer for solving differential equations.
Digital Equipment
Corporation introduces the
first minicomputer, the PDP-1,
to be followed with the PDP-8
in 1965.
1964
IBM’s 360 mainframe
becomes the standard
computer for major businesses.
1965
John Kemeny and
Thomas Kurtz
(both American)
develop the BASIC
computer language.
PRINT
FOR Counter = 1TO Items
PRINT USING “##.”; Counter;
LOCATE , ItemColumn
PRINT Item$(Counter);
LOCATE , PriceColumn
PRINT Price$(Counter)
NEXT Counter
1-1 HISTORICAL TIMELINE
9
Chronology 1-3: TIMELINE FOR COMPUTER TECHNOLOGY (continued)
Computer Technology
1968
Douglas Engelbart (American) demonstrates a
word-processor system, the mouse pointing device
and the use of “windows.”
1971
Texas Instruments introduces the pocket
calculator.
1971
Ted Hoff (American) invents the Intel
4004, which is the first computer microprocessor.
1976
IBM introduces the laser printer.
1976
Apple Computer sells
Apple I in kit form, which
is followed by the fully
assembled Apple II in
1977 and the Macintosh
in 1984.
1980
Microsoft introduces the
MS-DOS computer disk
operating system.
Microsoft Windows
is marketed in 1985.
1981
1989
Tim Berners-Lee (British) invents the World Wide Web by
introducing a networked hypertext system.
1991
The internet connects to 600,000 hosts in more than 100
countries.
1995
Sun Microsystems introduces the Java programming
language.
1996
Sabeer Bhatia (Indian-American) and Jack Smith
(American) launch Hotmail, which is the first
webmail service.
1997
IBM’s Deep Blue computer defeats World Chess
Champion Garry Kasparov.
2002
The billionth personal computer is sold; the second
billion is reached in 2007.
2010
The iPad is introduced in 2010.
IBM introduces the PC.
10
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Astronomy:
The Very Large
Array of Radio
Telescopes
Global Positioning System (GPS)
Motor
LCD
Screen
Plasma
propulsion
Radar
Telecommunication
Optical fiber
Ultrasound transducer
Cell phone
Ablation catheter
Liver
Ultrasound
image
Microwave ablation for liver cancer treatment
Figure 1-2 Electromagnetics is at the heart of numerous systems and applications.
1-3 THE NATURE OF ELECTROMAGNETISM
11
Table 1-1 Fundamental SI units.
Dimension
Unit
Length
Mass
Time
Electric charge
Temperature
Amount of substance
Luminous intensity
meter
kilogram
second
coulomb
kelvin
mole
candela
Table 1-2 Multiple and submultiple prefixes.
Symbol
Prefix
Symbol
m
kg
s
C
K
mol
cd
exa
peta
tera
giga
mega
kilo
E
P
T
G
M
k
1018
1015
1012
109
106
103
milli
micro
nano
pico
femto
atto
m
µ
n
p
f
a
10−3
10−6
10−9
10−12
10−15
10−18
special sections called Technology Briefs that are scattered
throughout this book.
Magnitude
1-2 Dimensions, Units, and Notation
The International System of Units, abbreviated SI after its
French name Système Internationale, is the standard system
used in today’s scientific literature for expressing the units
of physical quantities. Length is a dimension and meter is
the unit by which it is expressed relative to a reference
standard. The SI system is based on the units for the seven
fundamental dimensions listed in Table 1-1. The units for
all other dimensions are regarded as secondary because they
are based on, and can be expressed in terms of, the seven
fundamental units. Appendix A contains a list of quantities
used in this book, together with their symbols and units.
For quantities ranging in value between 10−18 and 1018, a
set of prefixes arranged in steps of 103 are commonly used to
denote multiples and submultiples of units. These prefixes, all
of which were derived from Greek, Latin, Spanish, and Danish
terms, are listed in Table 1-2. A length of 5 × 10−9 m, for
example, may be written as 5 nm.
In EM, we work with scalar and vector quantities. In
this book, we use a medium-weight italic font for symbols
denoting scalar quantities, such as R for resistance, and a
boldface roman font for symbols denoting vectors, such as E
for the electric field vector. A vector consists of a magnitude
(scalar) and a direction, with the direction usually denoted by
a unit vector. For example,
E = x̂E,
(1.1)
where E is the magnitude of E and x̂ is its direction. A symbol
denoting a unit vector is printed in boldface with a circumflex
(ˆ) above it.
Throughout this book, we make extensive use of phasor
representation in solving problems involving electromagnetic
quantities that vary sinusoidally in time. Letters denoting
phasor quantities are printed with a tilde (∼) over the letter.
e is the phasor electric field vector corresponding to
Thus, E
the instantaneous electric field vector E(t). This notation is
discussed in more detail in Section 1-7.
Notation Summary
• Scalar quantity: medium-weight italic, such as C for
capacitance.
• Units: medium-weight roman, as in V/m for volts per
meter.
• Vector quantities: boldface roman, such as E for
electric field vector
• Unit vectors: boldface roman with circumflex (ˆ)
over the letter, as in x̂.
• Phasors: a tilde (∼) over the letter; Ee is the phasor
counterpart of the sinusoidally time-varying scalar
e is the phasor counterpart of the
field E(t), and E
sinusoidally time-varying vector field E(t).
1-3 The Nature of Electromagnetism
Our physical universe is governed by four fundamental forces
of nature:
• The nuclear force, which is the strongest of the four, but
its range is limited to subatomic scales, such as nuclei.
• The electromagnetic force exists between all charged
12
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
particles. It is the dominant force in microscopic systems,
such as atoms and molecules, and its strength is on the
order of 10−2 that of the nuclear force.
Fg21
m2
• The weak-interaction force, whose strength is only 10−14
that of the nuclear force. Its primary role is in interactions
involving certain radioactive elementary particles.
Fg12
ˆ 12
R
• The gravitational force is the weakest of all four forces,
having a strength on the order of 10−41 of the nuclear
force. However, it often is the dominant force in macroscopic systems, such as the solar system.
m1
R12
Figure 1-3 Gravitational forces between two masses.
This book focuses on the electromagnetic force and its consequences. Even though the electromagnetic force operates at the
atomic scale, its effects can be transmitted in the form of electromagnetic waves that can propagate through both free space
and material media. The purpose of this section is to provide an
overview of the basic framework of electromagnetism, which
consists of certain fundamental laws governing the electric and
magnetic fields induced by static and moving electric charges,
the relations between the electric and magnetic fields, and how
these fields interact with matter. As a precursor, however, we
will take advantage of our familiarity with gravitational force
by describing some of its properties because they provide a
useful analogue to those of electromagnetic force.
1-3.1 Gravitational Force: A Useful Analogue
According to Newton’s law of gravity, the gravitational
force Fg21 acting on mass m2 due to a mass m1 at a distance
R12 from m2 (Fig. 1-3) is given by
Fg21 = −R̂12
Gm1 m2
R212
(N),
(1.2)
where G is the universal gravitational constant, R̂12 is a unit
vector that points from m1 to m2 , and the unit for force
is newton (N). The negative sign in Eq. (1.2) accounts for
the fact that the gravitational force is attractive. Conversely,
Fg12 = −Fg21 , where Fg12 is the force acting on mass m1 due to
the gravitational pull of mass m2 . Note that the first subscript
of Fg denotes the mass experiencing the force and the second
subscript denotes the source of the force.
◮ The force of gravitation acts at a distance. ◭
The two objects do not have to be in direct contact for each to
experience the pull by the other. This phenomenon of action
at a distance has led to the concept of fields. An object of
mass m1 induces a gravitational field ψ 1 (Fig. 1-4) that does
not physically emanate from the object, yet its influence exists
at every point in space such that, if another object of mass m2
were to exist at a distance R12 from the object of mass m1 , the
object of mass m2 would experience a force acting on it equal
to
Fg21 = ψ 1 m2 ,
(1.3)
where
ψ 1 = −R̂
Gm1
R2
(N/kg).
(1.4)
In Eq. (1.4), R̂ is a unit vector that points in the radial direction
away from object m1 ; therefore, −R̂ points toward m1 . The
force due to ψ 1 acting on a mass m2 , for example, is obtained
from the combination of Eqs. (1.3) and (1.4) with R = R12 and
R̂ = R̂12 . The field concept may be generalized by defining the
gravitational field ψ at any point in space such that, when a
test mass m is placed at that point, the force Fg acting on it is
related to ψ by
Fg
.
(1.5)
ψ=
m
−Rˆ
m1
Gravitational
field ψ1
Figure 1-4 Gravitational field ψ 1 induced by a mass m1 .
1-3 THE NATURE OF ELECTROMAGNETISM
13
The force Fg may be due to a single mass or a collection of
many masses.
Fe21
Force on charge q1
due to charge q2
1-3.2 Electric Fields
The electromagnetic force consists of an electrical component
Fe and a magnetic component Fm .
+q2
ˆ 12
R
Fe12
+q1
R12
◮ The electrical force Fe is similar to the gravitational
force, but with two major differences:
(1) the source of the electrical field is electric charge, not
mass, and
(2) even though both types of fields vary inversely
as the square of the distance from their respective sources, electric charges may have positive or
negative polarity, resulting in a force that may be
attractive or repulsive. ◭
We know from atomic physics that all matter contains a
mixture of neutrons, positively charged protons, and negatively charged electrons with the fundamental quantity of
charge being that of a single electron, usually denoted by
the letter e. The unit by which electric charge is measured is
the coulomb (C), named in honor of the eighteenth-century
French scientist Charles Augustin de Coulomb (1736–1806).
The magnitude of e is
e = 1.6 × 10−19
(C).
Figure 1-5 Electric forces on two positive point charges in free
space.
where Fe21 is the electrical force acting on charge q2 due
to charge q1 when both are in free space (vacuum), R12 is
the distance between the two charges, R̂12 is a unit vector
pointing from charge q1 to charge q2 (Fig. 1-5), and ε0 is
a universal constant called the electrical permittivity of free
space [ε0 = 8.854 × 10−12 farad per meter (F/m)]. The two
charges are assumed to be isolated from all other charges.
The force Fe12 acting on charge q1 due to charge q2 is
equal to force Fe21 in magnitude, but opposite in direction:
Fe12 = −Fe21 .
The expression given by Eq. (1.7) for the electrical force
is analogous to that given by Eq. (1.2) for the gravitational
force, and we can extend the analogy further by defining the
existence of an electric field intensity E due to any charge q as
(1.6)
E = R̂
The charge of a single electron is qe = −e and that of a proton
is equal in magnitude but opposite in polarity: qp = e.
◮ Coulomb’s experiments demonstrated that:
(1) two like charges repel one another, whereas two
charges of opposite polarity attract,
(2) the force acts along the line joining the charges, and
(3) its strength is proportional to the product of the magnitudes of the two charges and inversely proportional
to the square of the distance between them. ◭
q
4πε0 R2
(V/m) (in free space),
(1.8)
where R is the distance between the charge and the observation
point, and R̂ is the radial unit vector pointing away from the
charge. Figure 1-6 depicts the electric field lines due to a
positive charge. For reasons that will become apparent in later
chapters, the unit for E is volt per meter (V/m).
◮ If a point charge q′ is present in an electric field E (due
to other charges), the point charge will experience a force
acting on it equal to Fe = q′ E. ◭
Electric charge exhibits two important properties.
These properties constitute what today is called Coulomb’s
law, which can be expressed mathematically as
Fe21 = R̂12
q1 q2
4πε0 R212
(N) (in free space),
(1.7)
◮ The first property of electric charge is encapsulated by
the law of conservation of electric charge, which states
that the (net) electric charge can neither be created nor
destroyed. ◭
− +
+ −
+
+ −
−
+ −
−
+
+
+ −
−
+
+
+ −
− +
+ −
− +
− +
− +
− +
−
−
+
+
−
−
+
−
−
− +
−
+
− +
+
+ −
− +
−
+ −
+
+
+ −
q
− +
−
+ −
+
+
+ −
+
+ −
+ −
−
+ −
+ −
−
−
+
+ −
+ −
−
+ −
+
+
+ −
−
− +
−
+
+
− +
− +
+
+ −
−
−
− +
+ −
−
+ −
−
+
Electric
field lines
+
−
+q
−
+
+
ˆ
R
− +
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
+ −
14
Figure 1-6 Electric field E due to charge q.
Figure 1-7 Polarization of the atoms of a dielectric material by
a positive charge q.
If a volume contains np protons and ne electrons, then its total
charge is
q = np e − nee = (np − ne )e
(C).
(1.9)
Even if some of the protons were to combine with an equal
number of electrons to produce neutrons or other elementary
particles, the net charge q remains unchanged. In matter,
the quantum mechanical laws governing the behavior of the
protons inside the atom’s nucleus and the electrons outside it
do not allow them to combine.
◮ The second important property of electric charge is
embodied by the principle of linear superposition, which
states that the total vector electric field at a point in space
due to a system of point charges is equal to the vector
sum of the electric fields at that point due to the individual
charges. ◭
This seemingly simple concept allows us in future chapters
to compute the electric field due to complex distributions of
charge without having to be concerned with the forces acting
on each individual charge due to the fields by all of the other
charges.
The expression given by Eq. (1.8) describes the field
induced by an electric charge residing in free space. Let us
now consider what happens when we place a positive point
charge in a material composed of atoms. In the absence of
the point charge, the material is electrically neutral with each
atom having a positively charged nucleus surrounded by a
cloud of electrons of equal but opposite polarity. Hence, at
any point in the material not occupied by an atom, the electric
field E is zero. Upon placing a point charge in the material, as
shown in Fig. 1-7, the atoms experience forces that cause them
to become distorted. The center of symmetry of the electron
cloud is altered with respect to the nucleus with one pole of
the atom becoming positively charged relative to the other
pole. Such a polarized atom is called an electric dipole, and
the distortion process is called polarization. The degree of
polarization depends on the distance between the atom and
the isolated point charge, and the orientation of the dipole is
such that the axis connecting its two poles is directed toward
the point charge, as illustrated schematically in Fig. 1-7. The
net result of this polarization process is that the electric fields
of the dipoles of the atoms (or molecules) tend to counteract
the field due to the point charge. Consequently, the electric
field at any point in the material is different from the field that
would have been induced by the point charge in the absence
of the material. To extend Eq. (1.8) from the free-space case
to any medium, we replace the permittivity of free space ε0
with ε , where ε is the permittivity of the material in which the
electric field is measured and is therefore characteristic of that
particular material. Thus,
q
(V/m).
4πε R2
(material with permittivity ε )
E = R̂
(1.10)
Often, ε is expressed in the form
ε = εr ε0
(F/m),
(1.11)
where εr is a dimensionless quantity called the relative permittivity or dielectric constant of the material. For a vacuum,
εr = 1; for air near the Earth’s surface, εr = 1.0006; and the
values of εr for materials that we have occasion to use in this
book are tabulated in Appendix B.
1-3 THE NATURE OF ELECTROMAGNETISM
15
In addition to the electric field intensity E, we often find it
convenient to also use a related quantity called the electric flux
density D given by
D = εE
(C/m2 ),
N
(1.12)
with unit of coulomb per square meter (C/m2 ).
B
◮ These two electric quantities, E and D, constitute
one of two fundamental pairs of electromagnetic fields.
The second pair consists of the magnetic fields discussed
next. ◭
Magnetic
field lines
S
1-3.3 Magnetic Fields
As early as 800 BC, the Greeks discovered that certain kinds
of stones exhibit a force that attracts pieces of iron. These
stones are now called magnetite (Fe3 O4 ), and the phenomenon
they exhibit is known as magnetism. In the thirteenth century,
French scientists discovered that, when a needle was placed on
the surface of a spherical natural magnet, the needle oriented
itself along different directions for different locations on the
magnet. By mapping the directions indicated by the needle, it
was determined that the magnetic force formed magnetic field
lines that encircled the sphere and appeared to pass through
two points diametrically opposite to each other. These points,
called the north and south poles of the magnet, were found to
exist for every magnet, regardless of its shape. The magneticfield pattern of a bar magnet is displayed in Fig. 1-8. It was
also observed that like poles of different magnets repel each
other and unlike poles attract each other.
◮ The attraction–repulsion property for magnets is similar
to the electric force between electric charges, except
for one important difference: Electric charges can be
isolated, but magnetic poles always exist in pairs. ◭
If a permanent magnet is cut into small pieces, no matter how
small each piece is, it will always have a north and a south
pole.
The magnetic lines surrounding a magnet represent the
magnetic flux density B. A magnetic field not only exists
around permanent magnets but also can be created by electric
current. This connection between electricity and magnetism
was discovered in 1819 by the Danish scientist Hans Oersted
(1777–1851), who observed that an electric current in a wire
caused a compass needle placed in its vicinity to deflect
and that the needle turned so that its direction was always
Figure 1-8 Pattern of magnetic field lines around a bar magnet.
z
I
φ̂
B
B
y
r
B
B
x
B
B
B
B
Figure 1-9 The magnetic field induced by a steady current
flowing in the z direction.
perpendicular to the wire and to the radial line connecting the
wire to the needle. From these observations, he deduced that
the current-carrying wire induced a magnetic field that formed
closed circular loops around the wire (Fig. 1-9). Shortly after
Oersted’s discovery, French scientists Jean-Baptiste Biot and
Felix Savart developed an expression that relates the magnetic
flux density B at a point in space to the current I in the
conductor. Application of their formulation, known today as
the Biot–Savart law, to the situation depicted in Fig. 1-9 for
16
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
a very long wire residing in free space leads to the result that
the magnetic flux density B induced by a constant current I
flowing in the z direction is given by
B = φ̂φ
µ0 I
2π r
(T),
(1.13)
where r is the radial distance from the current and φ̂φ is an
azimuthal unit vector expressing the fact that the magnetic
field direction is tangential to the circle surrounding the current. The magnetic field is measured in tesla (T), named in
honor of Nikola Tesla (1856–1943). Tesla was an electrical
engineer whose work on transformers made it possible to
transport electricity over long wires without too much loss.
The quantity µ0 is called the magnetic permeability of free
space [µ0 = 4π × 10−7 henry per meter (H/m)], and it is
analogous to the electric permittivity ε0 . In fact, as we will
see in Chapter 2, the product of ε0 and µ0 specifies c, which is
the velocity of light in free space:
c= √
1
= 3 × 108
µ0 ε0
(m/s).
◮ We stated earlier that E and D constitute one of two
pairs of electromagnetic field quantities. The second pair
is B and the magnetic field intensity H, which are related
to each other through µ :
(1.14)
B = µ H.
1-3.4
(1.16)
◭
Static and Dynamic Fields
In EM, the time variable t, or more precisely if and how
electric and magnetic quantities vary with time, is of crucial
importance. Before we elaborate further on the significance of
this statement, it will prove useful to define the following timerelated adjectives unambiguously:
• Static—describes a quantity that does not change with
time. The term dc (i.e., direct current) is often used as a
synonym for static to describe not only currents, but other
electromagnetic quantities as well.
• Dynamic—refers to a quantity that does vary with time,
but conveys no specific information about the character of
the variation.
• Waveform—refers to a plot of the magnitude profile of a
quantity as a function of time.
We noted in Section 1-3.2 that, when an electric charge q′ is
subjected to an electric field E, it experiences an electric force
Fe = q′ E. Similarly, if a charge q′ resides in the presence of a
magnetic flux density B, it experiences a magnetic force Fm ,
but only if the charge is in motion and its velocity u is in
a direction not parallel (or anti-parallel) to B. In fact, as we
learn in more detail in Chapter 5, Fm points in a direction
perpendicular to both B and u.
To extend Eq. (1.13) to a medium other than free space, µ0
should be replaced with µ , which is the magnetic permeability
of the material in which B is being observed. The majority of
natural materials are nonmagnetic, meaning that they exhibit
a magnetic permeability µ = µ0 . For ferromagnetic materials,
such as iron and nickel, µ can be much larger than µ0 . The
magnetic permeability µ accounts for magnetization properties of a material. In analogy with Eq. (1.11), µ of a particular
material can be defined as
µ = µr µ0
(H/m),
(1.15)
where µr is a dimensionless quantity called the relative magnetic permeability of the material. The values of µr for commonly used ferromagnetic materials are given in Appendix B.
• Periodic—a quantity is periodic if its waveform repeats
itself at a regular interval, namely its period T . Examples
include the sinusoid and the square wave. By application
of the Fourier series analysis technique, any periodic
waveform can be expressed as the sum of an infinite series
of sinusoids.
• Sinusoidal—also called ac (i.e., alternating current),
describes a quantity that varies sinusoidally (or cosinusoidally) with time.
In view of these terms, let us now examine the relationship
between the electric field E and the magnetic flux density B.
Because E is governed by the charge q and B is governed by
I = dq/dt, one might expect that E and B must be somehow
related to each other. As we learn later, they may or may not
be interrelated, depending on whether I is static or dynamic.
Let us start by examining the dc case in which I remains
constant with time. Consider a small section of a beam of
charged particles, all moving at a constant velocity. The
moving charges constitute a dc current. The electric field due
to that section of the beam is determined by the total charge q
contained in it. The magnetic field does not depend on q,
but rather on the rate of charge (current) flowing through that
section. Few charges moving very fast can constitute the same
1-3 THE NATURE OF ELECTROMAGNETISM
17
Table 1-3 The three branches of electromagnetics.
Branch
Condition
Field Quantities (Units)
Electrostatics
Stationary charges
(∂ q/∂ t = 0)
Electric field intensity E (V/m)
Electric flux density D (C/m2 )
D = εE
Magnetostatics
Steady currents
(∂ I/∂ t = 0)
Magnetic flux density B (T)
Magnetic field intensity H (A/m)
B = µH
Dynamics
(time-varying fields)
Time-varying currents
(∂ I/∂ t 6= 0)
E, D, B, and H
(E, D) coupled to (B, H)
current as many charges moving slowly. In these two cases, the
induced magnetic field is the same because the current I is the
same, but the induced electric field is quite different because
the numbers of charges are not the same.
Electrostatics and magnetostatics refer to the study of EM
under the specific, respective conditions of stationary charges
and dc currents. They represent two independent branches, so
characterized because the induced electric and magnetic fields
do not couple to each other. Dynamics, the third and more
general branch of electromagnetics, involves time-varying
fields induced by time-varying sources, that is, currents and
associated charge densities. If the current associated with the
beam of moving charged particles varies with time, then the
amount of charge present in a given section of the beam also
varies with time, and vice versa. As we will see in Chapter 6,
the electric and magnetic fields become coupled to each other
in that case.
◮ A time-varying electric field generates a time-varying
magnetic field, and vice versa. ◭
Table 1-3 provides a summary of the three branches of
electromagnetics.
The electric and magnetic properties of materials are characterized by the parameters ε and µ , respectively. A third
fundamental parameter also needed is the conductivity of a
material σ , which is measured in siemens per meter (S/m).
The conductivity characterizes the ease with which charges
(electrons) can move freely in a material. If σ = 0, the charges
do not move more than atomic distances, and the material
is said to be a perfect dielectric. Conversely, if σ = ∞, the
charges can move very freely throughout the material, which
is then called a perfect conductor.
Table 1-4 Constitutive parameters of materials.
Parameter
Units
Free-Space Value
Electrical
permittivity ε
F/m
ε0 = 8.854 × 10−12
≈
1
× 10−9
36π
Magnetic
permeability µ
H/m
µ0 = 4π × 10−7
Conductivity σ
S/m
0
◮ The parameters ε , µ , and σ are often referred to as
the constitutive parameters of a material (Table 1-4).
A medium is said to be homogeneous if its constitutive
parameters are constant throughout the medium. ◭
What are the four fundamental
forces of nature, and what are their relative strengths?
Concept Question 1-1:
Concept Question 1-2:
What is Coulomb’s law? State
its properties.
What are the two important
properties of electric charge?
Concept Question 1-3:
(continued on p. 22)
18
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Technology Brief 1: LED Lighting
After lighting our homes, buildings, and streets for
over 100 years, the incandescent light bulb created
by Thomas Edison (1879) will soon become a relic of
the past. Many countries have taken steps to phase it
out and replace it with a much more energy-efficient
alternative: the light-emitting diode (LED).
(a)
Light Sources
The three dominant sources of electric light are
the incandescent, fluorescent, and LED light bulbs
(Fig. TF1-1). We examine each briefly.
(b)
Incandescent Light Bulb
◮ Incandescence is the emission of light from a hot
object due to its temperature. ◭
By passing electric current through a thin tungsten filament, which basically is a resistor, the filament’s temperature rises to a very high level, causing the filament to
glow and emit visible light. The intensity and shape of
the emitted spectrum depends on the filament’s temperature. A typical example is shown by the green curve in
Fig. TF1-2. The tungsten spectrum is similar in shape to
that of sunlight (yellow curve in Fig. TF1-2), particularly
in the blue and green parts of the spectrum (400–
550 nm). Despite the relatively strong (compared with
sunlight) yellow light emitted by incandescent sources,
the quasi-white light they produce has a quality that the
human eye finds rather comfortable.
(c)
Figure TF1-1 (a) Incandescent light bulb; (b) fluorescent
mercury vapor lamp; (c) white LED.
◮ The incandescent light
expensive to manufacture
LED light bulbs, but it
gard to energy efficacy
(Fig. TF1-7). ◭
bulb is significantly less
than the fluorescent and
is far inferior with reand operational lifetime
Of the energy supplied to an incandescent light bulb,
only about 2% is converted into light, with the remainder
wasted as heat! In fact, the incandescent light bulb is
the weakest link in the overall conversion sequence from
coal to light (Fig. TF1-3).
Fluorescent Light Bulb
To fluoresce means to emit radiation in consequence to
incident radiation of a shorter wavelength. By passing a
stream of electrons between two electrodes at the ends
of a tube (Fig. TF1-1(b)) containing mercury gas (or the
noble gases neon, argon, and xenon) at very low pressure, the electrons collide with the mercury atoms, causing them to excite their own electrons to higher energy
1-3 THE NATURE OF ELECTROMAGNETISM
19
Light-Emitting Diode
Energy (arbitrary units)
100
The LED contained inside the polymer jacket in
Fig. TF1-1(c) is a p-n junction diode fabricated on a
semiconductor chip. When a voltage is applied in a
forward-biased direction across the diode (Fig. TF1-4),
current flows through the junction and some of the
streaming electrons are captured by positive charges
(holes). Associated with each electron-hole recombining
act is the release of energy in the form of a photon.
Noon sunlight
75
Incandescent
tungsten
50
White LED
(with phosphor)
25
0
0.4
Fluorescent
mercury
0.5
0.6
0.7
Wavelength (micrometers)
Figure TF1-2 Spectra of common sources of visible light.
levels. When the excited electrons return to the ground
state, they emit photons at specific wavelengths, mostly
in the ultraviolet part of the spectrum. Consequently, the
spectrum of a mercury lamp is concentrated into narrow
lines, as shown by the blue curve in Fig. TF1-2.
◮ To broaden the mercury spectrum into one that
resembles that of white light, the inside surface of the
fluorescent light tube is coated with phosphor particles [such as yttrium aluminum garnet (YAG) doped
with cerium]. The particles absorb the UV energy
and then reradiate it as a broad spectrum extending
from blue to red; hence the name fluorescent. ◭
Photon
I
Photon
p-type
n-type
Holes
Electrons
+
_
_
V
Figure TF1-4 Photons are emitted when electrons combine
with holes.
Coal
Power plant
E1 = 0.35
e
Transmission lines
E2 = 0.92
Light
E3 = 0.024
Overall efficiency for conversion of chemical energy to light energy is
E1 × E2 × E3 = 0.35 × 0.92 × 0.024 ═ 0.8%
Figure TF1-3 Lighting efficiency. (Source: National Research Council, 2009.)
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
3
100
Energy (arbitrary units)
Spectral power (arbitrary units)
20
2
1
0
400
75
White LED
(phosphor-based blue LED)
50
25
Blue LED
500
600
Wavelength (nm)
700
Figure TF1-5 The addition of spectra from three monochromatic LEDs.
◮ The wavelength of the emitted photon depends on
the diode’s semiconductor material. The materials
most commonly used are aluminum gallium arsenide
(AIGaAs) to generate red light, indium gallium nitride
(InGaN) to generate blue light, and aluminum gallium
phosphide (AIGaP) to generate green light. In each
case, the emitted energy is confined to a narrow
spectral band. ◭
0
0.4
0.5
0.6
0.7
Wavelength (micrometers)
Figure TF1-6 Phosphor-based white LED emission spectrum.
With the blue LED/phosphor conversion technique,
a blue LED is used with phosphor powder particles
suspended in the epoxy resin that encapsulates it. The
blue light emitted by the LED is absorbed by the phosphor particles and then reemitted as a broad spectrum
(Fig. TF1-6). To generate high-intensity light, several
LEDs are clustered into a single enclosure.
Comparison
Two basic techniques are available for generating
white light with LEDs: (a) RGB and (b) blue/conversion.
The RGB approach involves the use of three monochromatic LEDs whose primary colors (red, green, and blue)
are mixed to generate an approximation of a whitelight spectrum. An example is shown in Fig. TF1-5. The
advantage of this approach is that the relative intensities
of the three LEDs can be controlled independently,
thereby making it possible to “tune” the shape of the
overall spectrum to generate an esthetically pleasing
color of “white.” The major shortcoming of the RGB
technique is cost: manufacturing three LEDs instead of
just one.
◮ Luminous efficacy (LE) is a measure of how
much light in lumens is produced by a light source
for each watt of electricity consumed by it. ◭
Of the three types of light bulbs we discussed, the
incandescent light bulb is by far the most inefficient
and its useful lifespan is the shortest (Fig. TF1-7).
For a typical household scenario, the 10-year cost—
including electricity and replacement cost—is several
times smaller for the LED than for the alternatives.
1-3 THE NATURE OF ELECTROMAGNETISM
Parameter
Type of Light Bulb
Incandescent
Luminous Efficacy
(lumens/W)
21
Fluorescent
White LED
Circa 2010
Circa 2025
~12
~40
~70
~150
Useful Lifetime
(hours)
~1000
~20,000
~60,000
~100,000
Purchase Price
~$1.50
~$5
~$10
~$5
Estimated Cost
over 10 Years
~$410
~$110
~$100
~$40
Figure TF1-7 Even though the initial purchase price of a white LED is several times greater than that of the incandescent light bulb, the
total 10-year cost of using the LED is only one-fourth of the incandescent’s (in 2010) and is expected to decrease to one-tenth by 2025.
22
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
What do the electrical permittivity and magnetic permeability of a material account
for?
Concept Question 1-4:
Answer: (a) B = 0, (b) B = 2B1 . (See
EM
.)
1-4 Traveling Waves
What are the three branches
and associated conditions of electromagnetics?
Concept Question 1-5:
Exercise 1-1: Given charges q1 = 10 mC, q2 = −10 mC,
and q3 = 5 mC, all in free space, what is the direction of
the force acting on charge q3 ?
y
Waves are a natural consequence of many physical processes:
Waves manifest themselves as ripples on the surfaces of
oceans and lakes; sound waves constitute pressure disturbances that travel through air; mechanical waves modulate
stretched strings; and electromagnetic waves carry electric
and magnetic fields through free space and material media as
microwaves, light, and X-rays. All of these various types of
waves exhibit a number of common properties, including:
• Moving waves carry energy.
q3
2m
• Waves have velocity; it takes time for a wave to travel
from one point to another. Electromagnetic waves in a
vacuum travel at a speed of 3 × 108 m/s, and sound waves
in air travel at a speed approximately a million times
slower, specifically 330 m/s. Sound waves cannot travel
in a vacuum.
2m
q1
q2
x
2m
Figure E1.1
Answer: Along +x̂ direction. [See
(the “ EM ” symbol
refers to the book website: em8e.eecs.umich.edu).]
EM
Exercise 1-2: Two parallel, very long, wires carry cur-
rents I1 and I2 . The magnetic field due to current I1 alone
is B1 . What is the magnetic field due to both currents at a
point midway between the wires if
(a) I1 = I2 and both currents flow along the +ŷ direction?
(b) I1 = I2 , but I2 flows along the −ŷ direction?
I1
I2
B=?
Figure E1.2
• Many waves exhibit a property called linearity. Waves
that do not affect the passage of other waves are called
linear because they can pass right through each other.
The total of two linear waves is simply the sum of
the two waves as they would exist separately. Electromagnetic waves are linear, as are sound waves. When
two people speak to one another, the sound waves they
generate do not interact with one another but simply
pass through each other. Water waves are approximately
linear; the expanding circles of ripples caused by two
pebbles thrown into two locations on a lake’s surface do
not affect each other. Although the interaction of the two
circles may exhibit a complicated pattern, it is simply
the linear superposition of two independent expanding
circles.
Waves are of two types: transient waves caused by sudden
disturbances and continuous periodic waves generated by a
repetitive source. We encounter both types of waves in this
book, but most of our discussion deals with the propagation of
continuous waves that vary sinusoidally with time.
An essential feature of a propagating wave is that it is a
self-sustaining disturbance of the medium through which it
travels. If this disturbance varies as a function of one space
variable, such as the vertical displacement of the string shown
in Fig. 1-10, we call the wave one-dimensional. The vertical
displacement varies with time and with the location along
the length of the string. Even though the string rises up
into a second dimension, the wave is only one-dimensional
1-4 TRAVELING WAVES
23
they include plane waves, cylindrical waves, and spherical
waves. A plane wave is characterized by a disturbance that
at a given point in time has uniform properties across an
infinite plane perpendicular to its direction of propagation
(Fig. 1-11(b)). Similarly, for cylindrical and spherical waves,
the disturbances are uniform across cylindrical and spherical
surfaces (Figs. 1-11(b) and (c)).
In the material that follows, we examine some of the basic
properties of waves by developing mathematical formulations
that describe their functional dependence on time and space
variables. To keep the presentation simple, we limit our
discussion to sinusoidally varying waves whose disturbances
are functions of only one space variable, and we defer the
discussion of more complicated waves to later chapters.
u
1-4.1 Sinusoidal Waves in a Lossless Medium
Regardless of the mechanism responsible for generating them,
all linear waves can be described mathematically in common
terms.
Figure 1-10 A one-dimensional wave traveling on a string.
because the disturbance varies with only one space variable.
A two-dimensional wave propagates out across a surface,
like the ripples on a pond (Fig. 1-11(a)), and its disturbance
can be described by two space variables. By extension, a
three-dimensional wave propagates through a volume, and its
disturbance may be a function of all three space variables.
Three-dimensional waves may take on many different shapes;
◮ A medium is said to be lossless if it does not attenuate
the amplitude of the wave traveling within it or on its
surface. ◭
By way of an example, let us consider a wave traveling on
a lake’s surface, and let us assume for the time being that
frictional forces can be ignored, thereby allowing a wave
generated on the water’s surface to travel indefinitely with no
loss in energy. If y denotes the height of the water’s surface
Cylindrical wavefront
Two-dimensional wave
(a) Circular waves
Spherical wavefront
Plane wavefront
(b) Plane and cylindrical waves
(c) Spherical wave
Figure 1-11 Examples of two-dimensional and three-dimensional waves: (a) circular waves on a pond, (b) a plane light wave exciting a
cylindrical light wave through the use of a long narrow slit in an opaque screen, and (c) a sliced section of a spherical wave.
24
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
relative to the mean height (undisturbed condition) and x
denotes the distance of wave travel, the functional dependence
of y on time t and the spatial coordinate x has the general form
or
2π t 2π x
y(x,t) = A cos
+ φ0
−
T
λ
(m),
(1.17)
where A is the amplitude of the wave, T is its time period,
λ is its spatial wavelength, and φ0 is a reference phase. The
quantity y(x,t) also can be expressed in the form
y(x,t) = A cos φ (x,t)
where
φ (x,t) =
2π t 2π x
+ φ0
−
T
λ
(m),
2π 2π dx
−
= 0,
T
λ dt
n = 0, 1, 2, . . .
(1.24)
which gives the phase velocity up as
dx λ
=
dt
T
(m/s).
(1.25)
(1.19)
The plots in Fig. 1-12 show the variation of y(x,t) with x at
t = 0 and with t at x = 0. The wave pattern repeats itself at a
spatial period λ along x and at a temporal period T along t.
If we take time snapshots of the water’s surface, the height
profile y(x,t) would exhibit the sinusoidal patterns shown
in Fig. 1-13. All three profiles correspond to three different
values of t, and the spacing between peaks is equal to the
wavelength λ , even though the patterns are shifted relative to
one another because they correspond to different observation
times. Because the pattern advances along the +x direction at
progressively increasing values of t, y(x,t) is called a wave
traveling in the +x direction. If we track a given point on
the wave, such as the peak P, and follow it in time, we can
measure the phase velocity of the wave. At the peaks of the
wave pattern, the phase φ (x,t) is equal to zero or multiples
of 2π radians. Thus,
2π t 2π x
−
= 2nπ ,
T
λ
The apparent velocity of that fixed height is obtained by taking
the time derivative of Eq. (1.23), so
up =
(rad).
(1.23)
(1.18)
The angle φ (x,t) is called the phase of the wave, and it should
not be confused with the reference phase φ0 , which is constant
with respect to both time and space. Phase is measured by the
same units as angles, that is, radians (rad) or degrees, with 2π
radians = 360◦ .
Let us first analyze the simple case when φ0 = 0:
2π t 2π x
y(x,t) = A cos
−
(m).
(1.20)
T
λ
φ (x,t) =
y 2π t 2π x
0
= cos−1
−
= constant.
T
λ
A
(1.21)
Had we chosen any other fixed height of the wave, say y0 ,
and monitored its movement as a function of t and x, this
again would have been equivalent to setting the phase φ (x,t)
constant such that
2π t 2π x
y(x,t) = y0 = A cos
(1.22)
−
λ
T
◮ The phase velocity, also called the propagation velocity, is the velocity of the wave pattern as it moves across
the water’s surface. ◭
The water itself mostly moves up and down; when the wave
moves from one point to another, the water does not move
physically along with it.
y(x, 0)
At t = 0
A
0
λ
2
−A
3λ
2
λ
x
λ
(a) y(x, t) versus x at t = 0
y(0, t)
At x = 0
A
0
−A
T
2
3T
2
T
t
T
(b) y(x, t) versus t at x = 0
Figure 1-12 Plots of y(x,t) = A cos
of (a) x at t = 0 and (b) t at x = 0.
2π t
T
− 2λπ x as a function
1-4 TRAVELING WAVES
25
compact form as
y(x, 0)
P
A
−A
λ
2
x
3λ
2
λ
2π
y(x,t) = A cos 2π f t −
x
λ
= A cos(ω t − β x),
(1.28)
(wave moving along +x direction)
(a) t = 0
up
where ω is the angular velocity of the wave and β is its phase
constant (or wavenumber), defined as
y(x, T/4)
P
A
−A
ω = 2π f
λ
2
x
3λ
2
λ
(b) t = T/4
y(x, T/2)
β=
2π
λ
(rad/s),
(1.29a)
(rad/m).
(1.29b)
In terms of these two quantities,
up = f λ =
P
A
3λ
2
λ
λ
2
−A
x
ω
.
β
(1.30)
So far, we have examined the behavior of a wave traveling
in the +x direction. To describe a wave traveling in the −x
direction, we reverse the sign of x in Eq. (1.28):
(c) t = T/2
− 2λπ x as a function
of x at (a) t = 0, (b) t = T /4, and (c) t = T /2. Note that the wave
moves in the +x direction with a velocity up = λ /T .
Figure 1-13 Plots of y(x,t) = A cos
2π t
T
The frequency of a sinusoidal wave, f , is the reciprocal of
its time period T :
f=
1
T
(Hz).
(1.26)
y(x,t) = A cos(ω t + β x).
(1.31)
(wave moving along −x direction)
◮ The direction of wave propagation is easily determined
by inspecting the signs of the t and x terms in the
expression for the phase φ (x,t) given by Eq. (1.19): If one
of the signs is positive and the other is negative, then the
wave is traveling in the positive x direction, and if both
signs are positive or both are negative, then the wave is
traveling in the negative x direction. The constant phase
reference φ0 has no influence on either the speed or the
direction of wave propagation. ◭
Combining the preceding two equations yields
up = f λ
(m/s).
(1.27)
The wave frequency f is measured in cycles per second and
has been assigned the unit hertz (Hz), named in honor of the
German physicist Heinrich Hertz (1857–1894), who pioneered
the development of radio-wave instrumentation.
Using Eq. (1.26), Eq. (1.20) can be rewritten in a more
We now examine the role of the phase reference φ0 given
previously in Eq. (1.17). If φ0 is not zero, then Eq. (1.28)
should be written as
y(x,t) = A cos(ω t − β x + φ0).
(1.32)
A plot of y(x,t) as a function of x at a specified t or as
a function of t at a specified x is shifted in space or time,
respectively, relative to a plot with φ0 = 0 by an amount
26
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Reference wave (φ0 = 0)
Lags behind reference wave
y
Leads ahead of
reference wave
A
φ0 = −π/4
φ0 = π/4
T
2
t
3T
2
T
−A
Figure 1-14 Plots of y(0,t) = A cos [(2π t/T ) + φ0 ] for three different values of the reference phase φ0 .
proportional to φ0 . This is illustrated by the plots shown in
Fig. 1-14. We observe that when φ0 is positive, y(t) reaches
its peak value, or any other specified value, sooner than when
φ0 = 0. Thus, the wave with φ0 = π /4 is said to lead the wave
with φ0 = 0 by a phase lead of π /4; and similarly, the wave
with φ0 = −π /4 is said to lag the wave with φ0 = 0 by a phase
lag of π /4. A wave function with a negative φ0 takes longer to
reach a given value of y(t), such as its peak, than the zerophase reference function.
Answer: (a) A = 6 V, (b) λ = 4 cm, (c) f = 150 Hz.
(See
EM
.)
Exercise 1-4: The wave shown in red in Fig. E1.4 is given
by υ = 5 cos 2π t/8. Of the following four equations:
(1) υ = 5 cos(2π t/8 − π /4),
(2) υ = 5 cos(2π t/8 + π /4),
◮ When its value is positive, φ0 signifies a phase lead in
time, and when it is negative, it signifies a phase lag. ◭
(3) υ = −5 cos(2π t/8 − π /4),
(4) υ = 5 sin 2π t/8,
Exercise 1-3: Consider the red wave shown in Fig. E1.3.
(a) which equation applies to the green wave? (b) which
equation applies to the blue wave?
What is the wave’s (a) amplitude, (b) wavelength, and
(c) frequency given that its phase velocity is 6 m/s?
υ (volts)
5
υ (volts)
6
4
2
0
-2
-4
-6
0
t (s)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
x (cm)
1
2
3
4
5
6
7
8
9 10
-5
Figure E1.4
Figure E1.3
Answer: (a) #2, (b) #4. (See
EM
.)
1-4 TRAVELING WAVES
27
Module 1.1 Sinusoidal Waveforms Learn how the shape of the waveform is related to the amplitude, frequency, and
reference phase angle of a sinusoidal wave.
Exercise 1-5: The electric field of a traveling electromag-
netic wave is given by
E(z,t) = 10 cos(π × 107t + π z/15 + π /6)
(V/m).
Determine (a) the direction of wave propagation, (b) the
wave frequency f , (c) its wavelength λ , and (d) its phase
velocity up .
Answer: (a) −z direction, (b) f = 5 MHz, (c) λ = 30 m,
(d) up = 1.5 × 108 m/s. (See
EM
.)
1-4.2 Sinusoidal Waves in a Lossy Medium
If a wave is traveling in the x direction in a lossy medium,
its amplitude decreases as e−α x . This factor is called the
attenuation factor, and α is called the attenuation constant
of the medium and its unit is neper per meter (Np/m). Thus, in
general,
(1.33)
y(x,t) = Ae−α x cos(ω t − β x + φ0).
The wave amplitude is now Ae−α x , not just A. Figure 1-15
shows a plot of y(x,t) as a function of x at t = 0 for A = 10 m,
λ = 2 m, α = 0.2 Np/m, and φ0 = 0. Note that the envelope of
the wave pattern decreases as e−α x .
28
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
y(x)
y(x)
10 m
Wave envelope
10e-0.2x
5m
0
1
2
3
4
5
6
7
8
x (m)
-5 m
-10 m
Figure 1-15 Plot of y(x) = (10e−0.2x cos π x) meters. Note that the envelope is bounded between the curve given by 10e−0.2x and its
mirror image.
The real unit of α is (1/m); the neper (Np) part is a dimensionless, artificial adjective traditionally used as a reminder
that the unit (Np/m) refers to the attenuation constant of
the medium, α . A similar practice is applied to the phase
constant β by assigning it the unit (rad/m) instead of just (l/m).
How can you tell if a wave is
traveling in the positive x direction or the negative x
direction?
Concept Question 1-6:
How does the envelope of the
wave pattern vary with distance in (a) a lossless medium
and (b) a lossy medium?
Concept Question 1-7:
is 1 kHz, the velocity of sound in water is 1.5 km/s, the wave
amplitude is 10 N/m2 , and p(x,t) was observed to be at its
maximum value at t = 0 and x = 0.25 m. Treat water as a
lossless medium.
Solution: According to the general form given by Eq. (1.17)
for a wave traveling in the positive x direction,
2π
2π
t−
p(x,t) = A cos
(N/m2 ).
x + φ0
T
λ
The amplitude A = 10 N/m2 , T = 1/ f = 10−3 s, and from
up = f λ ,
up
1.5 × 103
λ=
= 1.5 m.
=
f
103
Hence,
Concept Question 1-8:
Why does a negative value
of φ0 signify a phase lag?
Example 1-1:
Sound Wave in Water
An acoustic wave traveling in the x direction in a fluid (liquid
or gas) is characterized by a differential pressure p(x,t). The
unit for pressure is newton per square meter (N/m2 ). Find an
expression for p(x,t) for a sinusoidal sound wave traveling in
the positive x direction in water given that the wave frequency
4π
p(x,t) = 10 cos 2π × 103t − x + φ0
3
(N/m2 ).
Since at t = 0 and x = 0.25 m, p(0.25, 0) = 10 N/m2 , we have
−π
−4π
0.25 + φ0 = 10 cos
+ φ0 ,
10 = 10 cos
3
3
which yields the result (φ0 − π /3) = cos−1 (1), or φ0 = π /3.
Hence,
π
4π
3
(N/m2 ).
x+
p(x,t) = 10 cos 2π × 10 t −
3
3
1-4 TRAVELING WAVES
29
Module 1.2 Traveling Waves Learn how the shape of a traveling wave is related to its frequency and wavelength and to
the attenuation constant of the medium.
Example 1-2:
Solution: (a) Since the coefficients of t and x in the argument
of the cosine function have opposite signs, the wave must be
traveling in the +x direction.
Power Loss
(b)
A laser beam of light propagating through the atmosphere is
characterized by an electric field given by
E(x,t) = 150e−0.03x cos(3 × 1015t − 107x)
(V/m),
up =
3 × 1015
ω
=
= 3 × 108 m/s,
β
107
and this is equal to c, which is the velocity of light in free
space.
(c) At x = 200 m, the amplitude of E(x,t) is
where x is the distance from the source in meters. The attenuation is due to absorption by atmospheric gases. Determine
150e−0.03×200 = 0.37
(V/m).
(a) the direction of wave travel,
(b) the wave velocity, and
(c) the wave amplitude at a distance of 200 m.
Exercise 1-6: Consider the red wave shown in Fig. E1.6.
What is the wave’s (a) amplitude (at x = 0), (b) wavelength, and (c) attenuation constant?
30
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
electromagnetic spectrum (Fig. 1-16). Other members of this
family include gamma rays, X-rays, infrared waves, and radio
waves. Generically, they all are called EM waves because they
share the following fundamental properties:
υ (volts)
(2.8, 4.23)
5
(8.4, 3.02)
x (cm)
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
• The phase velocity of an EM wave propagating in a
vacuum is a universal constant given by the velocity of
light c, defined earlier by Eq. (1.14).
-5
Figure E1.6
Answer:
(See
EM
• A monochromatic (single frequency) EM wave consists
of electric and magnetic fields that oscillate at the same
frequency f .
(a) 5 V, (b) 5.6 cm, (c) α = 0.06 Np/cm.
• In a vacuum, the wavelength λ of an EM wave is related
to its oscillation frequency f by
.)
λ=
Exercise 1-7: The red wave shown in Fig. E1.7 is given
by υ = 5 cos 4π x (V). What expression is applicable to
(a) the blue wave and (b) the green wave?
υ (volts)
5V
3.52 V
1.01 V
5
0
x (m)
0.25
0.5
0.75
1.0
1.25
-5
Figure E1.7
Answer:
(a) υ = 5e−0.7x cos 4π x (V),
(b) υ = 5e−3.2x cos 4π x (V). (See
EM
.)
Exercise 1-8: An electromagnetic wave is propagating in
the z direction in a lossy medium with attenuation constant α = 0.5 Np/m. If the wave’s electric-field amplitude
is 100 V/m at z = 0, how far can the wave travel before
its amplitude is reduced to (a) 10 V/m, (b) 1 V/m, and
(c) 1 µ V/m?
Answer: (a) 4.6 m, (b) 9.2 m, (c) 37 m. (See
EM
.)
1-5 The Electromagnetic Spectrum
Visible light belongs to a family of waves arranged according
to frequency and wavelength along a continuum called the
c
.
f
(1.34)
Whereas all monochromatic EM waves share these properties,
each is distinguished by its own wavelength λ , or equivalently
by its own oscillation frequency f .
The visible part of the EM spectrum shown in Fig. 1-16
covers a very narrow wavelength range extending between
λ = 0.4 µ m (violet) and λ = 0.7 µ m (red). As we move
progressively toward shorter wavelengths, we encounter the
ultraviolet, X-ray, and gamma-ray bands, each so named
because of historical reasons associated with the discovery
of waves with those wavelengths. On the other side of the
visible spectrum lie the infrared band and then the microwave
part of the radio region. Because of the link between λ and
f given by Eq. (1.34), each of these spectral ranges may be
specified in terms of its wavelength range or its frequency
range. In practice, however, a wave is specified in terms of
its wavelength λ if λ < 1 mm, which encompasses all parts
of the EM spectrum except for the radio region, and the wave
is specified in terms of its frequency f if λ > 1 mm (i.e., in
the radio region). A wavelength of 1 mm corresponds to a
frequency of 3 × 1011 Hz = 300 GHz in free space.
The radio spectrum consists of several individual bands,
as shown in ‘the chart of Fig. 1-17 (see p. 32). Each band
covers one decade of the radio spectrum and has a letter
designation based on a nomenclature defined by the International Telecommunication Union. Waves of different frequencies have different applications because they are excited
by different mechanisms, and the properties of an EM wave
propagating in a nonvacuum material may vary considerably
from one band to another.
Although no precise definition exists for the extent of the
microwave band, it is conventionally regarded to cover the full
ranges of the UHF, SHF, and EHF bands. The EHF band is
sometimes referred to as the millimeter-wave band because
1-6 REVIEW OF COMPLEX NUMBERS
31
Optical window
Infrared windows
Radio window
Atmosphere opaque
100%
Ionosphere opaque
Atmospheric opacity
0
X-rays
Visible
Medical diagnosis
Ultraviolet
Infrared
Gamma rays
Cancer therapy
Heating,
Sterilization
night vision
1 fm
1 pm
-15
-12
10
1023
10
1021
1Å
-10
1 nm
10
10
1 EHz
-9
1 μm
10
1 PHz
1018
-6
1015
Radio spectrum
Communication, radar, radio and TV broadcasting,
radio astronomy
1 mm
1m
-3
1
1 km
10
1 THz
1 GHz
10
1 MHz
1012
109
106
3
1 Mm
10
1 kHz
6
Wavelength (m)
8
10
1 Hz
103
Frequency (Hz)
1
Figure 1-16 The electromagnetic spectrum.
the wavelength range covered by this band extends from 1 mm
(300 GHz) to 1 cm (30 GHz).
Concept Question 1-9:
What are the three fundamental
properties of EM waves?
is a useful shorthand representation for e jθ . Applying Euler’s
identity,
e jθ = cos θ + j sin θ ,
(1.38)
we can convert z from polar form, as in Eq. (1.37), into
rectangular form,
What is the range of frequencies covered by the microwave band?
Concept Question 1-10:
z = |z|e jθ = |z| cos θ + j|z| sin θ .
(1.39)
This leads to the relations
What is the wavelength range
of the visible spectrum? What are some of the applications
of the infrared band?
Concept Question 1-11:
1-6 Review of Complex Numbers
Any complex number z can be expressed in rectangular form
as
z = x + jy,
(1.35)
where x and y are the √
real (Re) and imaginary (Im) parts of z,
respectively, and j = −1. That is,
x = Re(z),
y = Im(z).
(1.36)
Alternatively, z may be cast in polar form as
z = |z|e jθ = |z| θ
x = |z| cos θ ,
p
|z| = + x2 + y2 ,
y = |z| sin θ ,
−1
θ = tan (y/x).
(1.40)
(1.41)
The two forms are illustrated graphically in Fig. 1-18. When
using Eq. (1.41), care should be taken to ensure that θ is
in the proper quadrant. Also note that, since |z| is a positive
quantity, only the positive root in Eq. (1.41) is applicable. This
is denoted by the + sign above the square-root sign.
The complex conjugate of z, denoted with a star superscript
(or asterisk), is obtained by replacing j (wherever it appears)
with − j, so that
z∗ = (x + jy)∗ = x − jy = |z|e− jθ = |z| −θ .
(1.42)
The magnitude |z| is equal to the positive square root of the
product of z and its complex conjugate:
(1.37)
where |z| is the magnitude of z, θ is its phase angle,and θ
|z| =
√
+
z z∗ .
(1.43)
32
TIO N A L TE LEC
O
M
M
300
E
285
RC
275
MOBILE
Maritime
Radionavigation
(Radio Beacons)
Aeronautical
Radionavigation
(Radio Beacons)
I N
M IN I
STR A T O
STANDARD
FREQUENCY
AND TIME SIGNAL
STANDARD
FREQUENCY
AND TIME SIGNAL
SATELLITE
Aeronautical
Mobile
AD
U.S. DEPARTMENT OF COMMERCE
MARITIME
RADIONAVIGATION
(RADIO BEACONS)
N
National Telecommunications and Information Administration
Office of Spectrum Management
AERONAUTICAL
RADIONAVIGATION
E
3000
ENT OF C O
RTM
MM
2850
AERONAUTICAL
MOBILE (R)
A
EP
29.7
29.8
29.89
29.91
30.0
200
October 2003
MOBILE
FIXED
AT
AMATEUR SATELLITE
LAND MOBILE
FIXED
FIXED
RM
AMATEUR
2502
2505
TI O NS & INF O
300
STANDARD FREQ. AND TIME SIGNAL
2501
U.S
.
IC
A
3000
Space Research
Aeronautical
Mobile
2900
Radiolocation
MARITIME
MOBILE
MARITIME
RADIONAVIGATION
2483.5
2500
2655
2690
2700
Radiolocation
190
AERONAUTICAL
RADIONAVIGATION
2495
STANDARD FREQ.
AERONAUTICAL
RADIONAVIGATION
FIXED
MOB** B- SAT. FX FX-SAT
RADIO ASTRON. SPACE RESEARCH EARTH EXPL SAT
AERONAUTICAL METEOROLOGICAL
AIDS
RADIONAVIGATION
IO
160
300 kHz
30.0
FX-SAT (S - E)
E-Expl Sat Radio Ast Space res.
2170
2173.5
2190.5
2194
MARITIME
MOBILE
MOBILE SATELLITE (E-S)
MOBILE SATELLITE (S-E)
BCST - SAT. MOBILE**
2107
MARITIME
MOBILE
STANDARD FREQ. AND TIME SIGNAL (2500kHz)
MOBILE
29.9
FIXED SATELLITE (E-S)
RADIODETERMINATION SAT. (S-E)
25.01
25.07
25.21
25.33
25.55
25.67
26.1
26.175
26.48
26.95
26.96
27.23
27.41
27.54
28.0
LAND
MOBILE
LAND MOBILE
29.5
24.89
24.99
25.005
3 MHz
27.5
MOBILE**
RADIO ASTRONOMY
BROADCASTING
MARITIME MOBILE
LAND MOBILE
MOBILE**
FIXED
FIXED
MOBILE**
FIXED
MOBILE**
LAND MOBILE
FIXED
FIXED
MOBILE
MOBILE
MARITIME MOBILE (TELEPHONY)
MOBILE (DISTRESS AND CALLING)
MARITIME MOBILE (TELEPHONY)
FIXED
25.25
25.5
27.0
MOBILE
FIXED
SAT (E-S)
2417
2450
30 MHz
INTERSATELLITE
MOBILE
Amateur
Radiolocation
MOBILE
LAND MOBILE
MARITIME MOBILE
LAND MOBILE
FIXED
ISM – 27.12 ± .163 MHz
FIXED
FIXED
AMATEUR
STANDARD FREQ. AND TIME SIGNAL (25,000 kHz)
Space Research
STANDARD FREQ.
2385
2390
2400
AMATEUR
FIXED
22.855
23.0
23.2
23.35
MOBILE**
AMATEUR SATELLITE
2360
Fixed
FIXED
AMATEUR
2065
MARITIME MOBILE (TELEPHONY)
MARITIME MOBILE
FIXED
Mobile*
FIXED
AERONAUTICAL MOBILE (OR)
235.0
R- LOC. B-SAT
RADIOLOCATION
MOBILE
FIXED
220.0
222.0
225.0
FIXED
2345
FX
MOB
Amateur
MOBILE
FIXED
FIXED
Earth Exploration
Satellite (S-S)
MOBILE
Radiolocation
25.05
FIXED SATELLITE (E-S) MOBILE SATELLITE (E-S)
Fixed
Radiolocation Mobile Fixed
300 MHz
Earth
std Exploration
freq e-e-sat
(S-S) (s-s)
&Satellite
time e-e-sat
SATELLITE
(E-S)
MOBILE
FIXED
FIXED
SATELLITE
(E-S)
INTER-SAT. MOBILE
INTER-SAT. MOBILE
MOBILE
24.75
MOBILE
Radiolocation
Radiolocation
MOBILE SATELLITE
INTER-SATELLITE
RADIONAVIGATION
Earth
Standard
Exploration Frequency and
Satellite
Time Signal
SATELLITE (E-S)
RADIOLOCATION
FIX
E D (E-S)
SATELLITE
3 GHz
MOBILE
30 GHz
FIXED
300 GHz
275.0
300.0
24.65
(S-S) F I X ESatellite
(E-S)
D
265.0
24.25
24.45
INTER-SATELLITE
RADIOLOCATION
FIXED
MOBILE
FIXED
RADIORADIONAVIGATION ASTRONOMY
SATELLITE
MOBILE
SATELLITE
MOBILE
FIXED
RADIONAVIGATION SATELLITE
(E-S)
252.0
Amateur
FIXED
RADIONAVIGATION
Earth
Expl.
INTER-SATELLITE
Satellite (Active)
FIXED
21.45
21.85
21.924
22.0
FIXED
250.0
BCST-SATELLITE
BROADCASTING
FIXED
AERONAUTICAL MOBILE (R)
216.0
Radiolocation
FIXED
110
130
2000
MARITIME
MOBILE
FIXED
MARITIME
MOBILE
248.0
Mobile
1900
90
RADIOLOCATION
MOBILE
MOBILE
241.0
Radiolocation
Mobile
Fixed
LAND MOBILE
AMATEUR
2290
2300
2305
2310
2320
19.68
19.80
19.990
19.995
20.005
20.010
LAND MOBILE
238.0
Radiolocation
ISM – 2450.0 ± 50 MHz
Radiolocation
Amateur
AMATEUR
EARTH EXPLORATION
SATELLITE (Passive)
MOBILE**
RADIOLOCATION MOBILE**
FIXED
Radiolocation Mobile Fixed MOB FX R- LOC. B-SAT
1800
FIXED
EARTH EXPL.
SAT. (Passive)
FIXED
SATELLITE (S-E)
Amateur Satellite
RADIOLOCATION
FIXED
Amateur
Amateur
1705
AMATEUR SATELLITE
AMATEUR
1605
1615
BROADCASTING
AMATEUR
21.0
2160
2180
2200
SPACE
EARTH
OPERATION EXPLORATION
(s-E)(s-s) SAT. (s-E)(s-s)
SPACE
RESEARCH
(s-E)(s-s)
MOBILE
(LOS)
SPACE RES..(S-E)
2155
17.9
17.97
18.03
18.068
18.168
18.78
18.9
19.02
BROADCASTING
MOBILE
Mobile
FIXED
MOBILE
FIXED
SPACE RES.
SATELLITE(S-E) (Passive)
MOBILE
Earth Expl.
Satellite
(Active)
ISM – 24.125 ± 0.125 GHz
ISM – 245.0 ± 1GHz
MOBILE
FIXED
RADIOLOCATION
AMATEUR SATELLITE
SPACE RES. (Passive)
FIXED
(LOS)
24.0
24.05
235.0
MOBILE
MOBILE SATELLITE (S-E)
23.6
EARTH EXPL.
SAT. (Passive)
AMATEUR SATELLITE
2025
2110
FIXED
EARTH
EXPLORATION
SATELLITE
(Passive)
SPACE
RESEARCH
(Passive)
RADIO
ASTRONOMY
AMATEUR
FIXED
MARITIME MOBILE
FIXED
STAND. FREQ. & TIME SIG.
Space Research
STANDARD FREQUENCY & TIME SIGNAL (20,000 KHZ)
Space Research
STANDARD FREQ.
17.41
17.48
17.55
D
FIXED
SATELLITE
(E-S)
MOBILE
FIXED
FIXED
Radiolocation
FIXED
23.55
SPACE RES.
(Passive)
RADIO ASTRONOMY
S)
FIXED
MOBILE
FIXED
FIXED
SATELLITE (S-E)
2020
MOB. FX.
MOBILE
INTER-SATELLITE
MOBILE
Land Mobile
MOBILE
1850
2000
MOBILE
SPACE OP.
(E-S)(s-s)
BROADCASTING
AERONAUTICAL MOBILE (R)
AERONAUTICAL MOBILE (OR)
FIXED
AMATEUR
AMATEUR SATELLITE
Mobile
FIXED
MARITIME MOBILE
BROADCASTING
FIXED
173.2
173.4
174.0
1755
MOBILE
MOBILE SATELLITE (E-S)
FIXED
FIXED
22.55
FIXED
231.0
MOBILE
MOBILE
SPACE RES. EARTH EXPL.
(E-S)(s-s)
SAT. (E-S)(s-s)
FIXED
217.0
FIXED
MOBILE
FIXED
FIXED
21.4
22.0
22.21
22.5
FIXED
FIXED
1710
FIXED
FIXED
BROADCASTING
FIXED
MOBILE
202.0
1700
16.36
MARITIME
MOBILE
FIXED
UN
200.0
19.3
19.7
20.1
20.2
21.2
1675
156.2475
157.0375
157.1875
157.45
161.575
161.625
161.775
162.0125
FIXED
EARTH
EXPLORATION SAT.
(Passive)
METEOROLOGICAL
AIDS (Radiosonde)
1670
61
70
MARITIME
MOBILE
RADIONAVIGATION
SATELLITE
SPACE RES.
(Passive)
FIXED
MOBILE**
METEOROLOGICAL
SATELLITE (s-E)
MARITIME MOBILE
LAND MOBILE
FIXED
LAND MOBILE
MARITIME MOBILE
LAND MOBILE
MARITIME MOBILE
LAND MOBILE
1660
1660.5
1668.4
15.6
15.8
MARITIME
MOBILE
MOBILE
SATELLITE
RADIONAVIGATION
MOBILE
MOBILE
METEOROLOGICAL
AIDS (RADIOSONDE)
BROADCASTING
(TV CHANNELS 7-13)
FIXED
FIXED SATELLITE (S-E)
MOBILE SAT. (S-E)
FIXED SATELLITE (S-E)
FX SAT (S-E)
MOBILE SATELLITE (S-E)
FXFIXED
SAT (S-E) MOBILE SAT (S-E)
STD FREQ. & TIME
SPACE RES.
F I X E D MOBILE EARTH EXPL. SAT.
FIXED
MOBILE
MOBILE**
FIXED
S P A C E R A D . A S T MOBILE** F I X E D EARTH EXPL. SAT.
RES.
190.0
FIXED
SPACE RESEARCH (Passive)
FIXED SATELLITE (S-E)
185.0
MOBILE SAT. (E-S)
RADIO ASTRONOMY
RADIO ASTRONOMY
RADIO ASTRONOMY
BROADCASTING
FIXED
182.0
17.2
17.3
17.7
17.8
18.3
18.6
18.8
FIXED
MOBILE
176.5
17.1
BROADCASTING
154.0
RADIOLOCATION
FIXED
FIXED SATELLITE (S-E)
FIXED
FIXED SATELLITE (S-E)
FX SAT (S-E) EARTH EXPL. SAT.
SPACE RES.
FIXED SATELLITE (E-S)
MARITIME MOBILE
MARITIME MOBILE
MOBILE SATELLITE (E-S)
15.63
15.7
16.6
Fixed
INTERSATELLITE
MOBILE
FX SAT (E-S)
Radioloc.
Radiolocation
FIXED
EARTH
SPACE RESEARCH
EXPLORATION
(Passive)
SATELLITE (Passive)
FIXED
Radiolocation
Radiolocation
Radiolocation
RADIOLOC.
MET. SAT.
(s-E)
RADIO
ASTRONOMY
BCST SAT.
INTERSATELLITE
MOBILE
FIXED
174.5
Space Res.
150.8
152.855
59
FIXED
EARTH
INTERSATELLITE EXPLORATION
SAT. (Passive)
Earth Expl Sat
LAND MOBILE
14.990
15.005
15.010
15.10
RADIONAVIGATION Radiolocation
SPACE
RESEARCH
(Passive)
170.0
LAND MOBILE
LAND MOBILE
FIXED
STANDARD FREQ. AND TIME SIGNAL (60 kHz)
Mobile*
STANDARD FREQ. AND TIME SIGNAL (15,000 kHz)
Space Research
STANDARD FREQ.
AERONAUTICAL MOBILE (OR)
MOBILE
MOBILE
FIXED SAT (E-S)
AERONAUTICAL RADIONAVIGATION
RADIOLOCATION
RADIOLOCATION Space Res.(act.)
RADIOLOCATION
168.0
FIXED
FIXED
RADIONAVIGATION
AERONAUTICAL RADIONAVIGATION
164.0
1558.5
1559
1610
1610.6
1613.8
1626.5
FIXED
FIXED
INTERSATELLITE
15.4
15.43
1549.5
144.0
146.0
148.0
149.9
150.05
FIXED
SPACE RES.
(Passive)
MOBILE
MOBILE
AERONAUTICAL MOBILE SATELLITE (R) (space to Earth)
AERONAUTICAL RADIONAVIGATION
RADIONAV. SATELLITE (Space to Earth)
AERO. RADIONAVIGATION RADIO DET. SAT. (E-S) M O B I L E S A T ( E - S )
AERO. RADIONAV. RADIO DET. SAT. (E-S) MOBILE SAT. (E-S) RADIO ASTRONOMY
Mobile Sat. (S-E)
AERO. RADIONAV. RADIO DET. SAT. (E-S) MOBILE SAT. (E-S)
15.1365
15.35
AMATEUR SATELLITE
AMATEUR
MOBILE
MOBILE SATELLITE (E-S)
FIXED
RADIONAV-SATELLITE
MOBILE SATELLITE (E-S)
FIXED
MOBILE
13.2
13.26
13.36
13.41
13.57
13.6
13.8
13.87
14.0
14.25
14.35
FIXED
Space Research
EARTH EXPL. SAT.
(Passive)
MOBILE SATELLITE
(Space to Earth)
MOBILE
AMATEUR
BROADCASTING
FIXED
BROADCASTING
Mobile*
FIXED
AMATEUR
AMATEUR SATELLITE
AMATEUR
MOBILE
SATELLITE
Space Research
Space Research
Mobile
SPACE RESEARCH
(Passive)
14.47
14.5
14.7145
FIXED
1544
137.0
137.025
137.175
137.825
138.0
FIXED
SATELLITE
Mobile
Fixed
FIXED
RADIO ASTRONOMY
14.4
1530
1535
1545
Mobile Satellite (S- E)
136.0
MET. SAT. (S-E)
MET. SAT. (S-E)
MET. SAT. (S-E)
MET. SAT. (S-E)
SPACE OPN. (S-E)
SPACE OPN. (S-E)
SPACE OPN. (S-E)
SPACE OPN. (S-E)
RADIONAVIGATION
SATELLITE
FIXED
Land Mobile
SAT. (E-S) Satellite (E-S)
FX SAT.(E-S) L M Sat(E-S)
MOBILE SATELLITE (S-E)
SPACE RES. (S-E)
SPACE RES. (S-E)
SPACE RES. (S-E)
SPACE RES. (S-E)
12.23
Mobile*
SPACE OPERATION
151.0
FIXED
SATELLITE
(S-E)
FIXED
RADIO
ASTRONOMY
FIXED
FIXED
Mobile **
AERONAUTICAL MOBILE SATELLITE (R)
(space to Earth)
AERONAUTICAL MOBILE SATELLITE (R)
(space to Earth)
MOB. SAT. (S-E)
Mob. Sat. (S-E)
MOB. SAT. (S-E)
Mob. Sat. (S-E)
132.0125
12.10
MARITIME MOBILE
150.0
Mobile
MOBILE SAT.
(Space to Earth)
Mobile
(Aero. TLM)
MOBILE SAT.
MARITIME MOBILE SAT.
(Space to Earth)
(Space to Earth)
MARITIME MOBILE SATELLITE
MOBILE SATELLITE (S-E)
(space to Earth)
Land Mobile
Satellite (E-S)
Mobile
AERO RADIONAV
EARTH
EXPLORATION
SATELLITE (Passive)
MOBILE (AERONAUTICAL TELEMETERING)
1430
1432
1435
1525
AERONAUTICAL MOBILE (R)
RADIOASTRONOMY
FIXED
BROADCASTING
FIXED
Radiolocation
Fixed
Fixed
FIXED
MOBILE
LAND MOBILE (TLM)
MOBILE
128.8125
AERONAUTICAL
MOBILE (R)
AERONAUTICAL MOBILE (R)
AERONAUTICAL MOBILE (R)
12.05
FIXED
MARITIME
MOBILE
AERONAUTICAL MOBILE (OR)
TRAVELERS INFORMATION STATIONS (G) AT 1610 kHz
EARTH
EXPL. SAT. SPACE RES.
(Passive)
(Passive)
MOBILE
149.0
FIXED (TLM)
(S-E)
121.9375
123.0875
123.5875
AERONAUTICAL
MOBILE (R)
SPACE RESEARCH
FIXED
SATELLITE (S-E)
Fixed (TLM)
FIXED (TLM)
1390
1392
1395
1400
1427
1429.5
MARITIME
MOBILE
MOBILE
FIXED
SATELLITE (E-S)
Mobile**
SPA CE RESEARCH ( Passive)
FIXED**
14.2
142.0
144.0
EARTH EXPL SAT (Passive)
FIXED
Amateur Satellite
FIXED-SAT
13.25
Space
Research
FIXED
SATELLITE
(S-E)
Amateur
ASTRONOMY
LAND MOBILE
LAND MOBILE (TLM)
FIXED
Space Research (E-S)
AERONAUTICAL RADIONAV.
13.4
RADIOStandard
RadioLOCATION
Freq. and
location
Time Signal
13.75
FIXED
RADIORadioSatellite (E-S)
SAT.(E-S) location
LOCATION
14.0
RADIO
Land Mobile
FIXED
Space
NAVIGATION
SAT. (E-S) Satellite (E-S)
Research
Radiolocation
MOBILE
SATELLITE
AMATEUR SATELLITE
FIXED
FIXED
RADIONAVIGATION
SATELLITE
RADIONAVIGATION
MOBILE
AMATEUR
(E-S)
FIXED
LAND MOBILE
12.75
FIXED
SATELLITE MOBILE
(E-S)
SPACE
RESEARCH (S-E)
(Deep Space)
RAD IOLOCATION
FIXED-SAT
MOBILE **
MOBILE **
RADIO
FIXED
MOBILE
ISM – 13.560 ± .007 MHz
INTERSATELLITE
RADIOLOCATION
FIXED
MOBILE
FIXED
SATELLITE (E-S)
MOBILE
FIXED
12.7
126.0
134.0
RADIOLOCATION
FIXED
FIXED
(Passive)
(Passive)
NA
BROADCASTING
SATELLITE
INTER-
BROADCASTING
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
SPACE EARTH
Amatuer
RES. EXPL . SAT
120.02
EARTH
SPACE
INTERSAT.
MOBILE SATELLITE RESEARCH EXPL SAT.
11.175
11.275
11.4
11.6
11.65
BROADCASTING
FIXED
RADIO SERVICES COLOR LEGEND
MOBILE
FIXED
AERONAUTICAL MOBILE (OR)
AERONAUTICAL MOBILE (R)
FIXED
BROADCASTING
FIXED
117.975
AERONAUTICAL
MOBILE (R)
AERONAUTICAL MOBILE
AERONAUTICAL MOBILE
1350
30
MARITIME
MOBILE
SATELLITE
MARITIME
RADIONAVIGATION
METEOROLOGICAL
AIDS
METEOROLOGICAL
SATELLITE
Radiolocation
1300
12.2
119.98
9.9
9.995
10.003
10.005
10.1
10.15
MARITIME MOBILE
FIXED
FIXED
SATELLITE
(S-E)
Mobile **
ISM – 122.5 ± .500 GHz
116.0
EARTH
SPACE
(Passive)
30
9.5
FIXED
AERONAUTICAL
RADIONAVIGATION
1240
11.7
INTERF I X E D MOBILE SATELLITE
RESEARCH EXPL SAT.
(Passive)
FIXED
STANDARD FREQ. AND TIME SIGNAL (10,000 kHz)
Space Research
STANDARD FREQ.
AERONAUTICAL MOBILE (R)
AMATEUR
AERONAUTICAL
RADIONAVIGATION
FIXED
SATELLITE
(S-E)
FIXED
EARTH
EXPLORATION
SATELLITE
(Passive)
SPACE
RESEARCH
(Passive)
RADIO
ASTRONOMY
Amateur
8.965
9.040
9.4
BROADCASTING
BROADCASTING
108.0
RADIOLOCATION
RADIOLOCATION
8.815
FIXED
1215
RADIONAVIGATION
SATELLITE (S-E)
20.05
8.1
8.195
FIXED
10.7
19.95
AMATEUR
SATELLITE
10.5
10.55
10.6
10.68
INTERSATELLITE
AERONAUTICAL MOBILE (OR)
10.45
FIXED
EARTH EXPL.
SATELLITE (Passive)
FREQUENCY
ALLOCATIONS
RADIO
ASTRONOMY
SPACE
RESEARCH (Passive)
AERONAUTICAL MOBILE (R)
Mobile*
RADIO
ASTRONOMY
MARITIME
MARITIMEMOBILE
MOBILE
FIXED
AERONAUTICAL
RADIONAVIGATION
Amateur
Satellite
Amateur
FIXED
EARTH EXPL.
SAT. (Passive)
SPACE RESEARCH
(Passive)
7.3
7.35
RADIO
ASTRONOMY
RADIODETERMINATION
SATELLITE
RADIOLOCATION
Amateur
Radiolocation
RADIOLOCATION
102.0
105.0
FIXED
FIXED
7.1
BROADCASTING
BROADCASTING
SATELLITE
EARTH
EXPLORATION
SATELLITE
RADIOLOCATION
Radiolocation
100.0
FIXED
FIXED
10.0
88.0
928
929
930
931
932
935
940
941
944
960
FIXED
FIXED
FIXED
FIXED
LAND MOBILE
LAND MOBILE
Radiolocation
7.0
STANDARD FREQ. AND TIME SIGNAL (20 kHz)
FIXED
FIXED
Amateur
FIXED
LAND MOBILE
MOBILE
LAND MOBILE
BROADCASTING
BROADCASTING
BROADCASTING
(AM RADIO)
9.5
FIXED
FIXED
RADIOLOCATION
UNITED
STATES
9.3
THE RADIO SPECTRUM
9.2
AERONAUTICAL MOBILE
MOBILE
FIXED
FIXED
Mobile
Mobile
74.6
74.8
75.2
75.4
76.0
RADIOLOCATION
SATELLITE
LAND MOBILE
LAND
MOBILE
LAND
MOBILE
SATELLITE
MARITIME
MOBILE
LAND MOBILE
FIXED
LAND MOBILE
LAND MOBILE
FIXED
8.45
9.0
821
824
849
851
866
869
894
896
901901
902
AERONAUTICAL MOBILE
8.4
8.215
Radiolocation
RADIOLOCATION
Radiolocation
RADIONAVIGATION
MOBILE
SATELLITE
RADIONAVIGATION
SATELLITE
MOBILE
SPACE RESEARCH
(Passive)
FIXED
SATELLITE
(S-E)
FIXED
FIXED
LAND MOBILE
LAND MOBILE
FIXED
8.175
95.0
EARTH EXPL.
SATELLITE (Passive)
MOBILE
FIXED
AMATEUR SATELLITE
AMATEUR
MARITIME MOBILE
Radiolocation
FIXED
MOBILE
FIXED
AERONAUTICAL RADIONAVIGATION
FIXED
MOBILE
FIXED
MOBILE
794
806
8.025
73.0
RADIO ASTRONOMY
BROADCAST
LAND MOBILE
7.75
Radiolocation
Meteorological
Aids
MOBILE
BROADCASTING
(FM RADIO)
FIXED
SATELLITE
(E-S)
RADIOLOCATION
MOBILE
FIXED
7.55
6.685
6.765
AERONAUTICAL MOBILE (OR)
Mobile
AMATEUR
72.0
MOBILE
FIXED
7.45
8.5
92.0
FIXED
764
FIXED
EARTH
EXPLORATION
SATELLITE
(Passive)
SPACE
RESEARCH
(Passive)
RADIO
ASTRONOMY
AERONAUTICAL
RADIONAVIGATION
MARITIME
RADIONAVIGATION
RADIONAVIGATION
BROADCAST
MOBILE
776
AERONAUTICAL
MOBILE
AERONAUTICAL
MOBILE
SATELLITE
AERONAUTICAL
RADIONAVIGATION
AMATEUR
MOBILE
Fixed
MOBILE
FIXED
FIXED
Mobile Satellite
(E-S)(no airborne)
FIXED
Radiolocation
14
MARITIME MOBILE
Mobile
Satellite (E-S)
(no airborne)
MET.
SATELLITE
(E-S)
FIXED
SPACE RESEARCH (S-E)
(deep space only)
RADIOLOCATION
RADIONAVIGATION
FIXED
SATELLITE
FIXED
(E-S)
FIXED
SATELLITE
(E-S)
FIXED
6.2
MARITIME MOBILE
BROADCAST
FIXED
FIXED
7.30
7.90
SPACE RESEARCH (S-E)
5.68
5.73
5.90
MARITIME
MOBILE
EARTH EXPL.
SATELLITE (S-E)
535
FIXED
EARTH EXPL.
SAT. (S-E)
9
525
5.95
698
7.19
7.235
7.25
Fixed
Mobile
Satellite (E-S)
AERONAUTICAL
RADIONAVIGATION
(RADIO BEACONS)
AERONAUTICAL MOBILE (R)
ISM – 915.0 ± 13 MHz
FIXED
FIXED
FIXED
MOBILE
SATELLITE (E-S)
SATELLITE (E-S)
FIXED
EARTH EXPL.
FIXED
SATELLITE (E-S) SATELLITE(S-E)
MOBILE
BROADCASTING
BROADCASTING
(TV CHANNELS 5-6)
FIXED
AMATEUR SATELLITE
76.0
Amateur
77.0
Amateur
Amateur Sat.
77.5
AMATEUR
AMATEUR SAT 78.0
Amateur
Amateur
Satellite
81.0
FIXED
MOBILE
MOBILE
SATELLITE SATELLITE
(S-E)
(S-E)
84.0
BROADBROADMOBILE
CASTING CASTING
SATELLITE
86.0
AERONAUTICAL
RADIONAVIGATION
(RADIO BEACONS)
BROADCASTING
Mobile
75.5
AMATEUR
RADIOLOC.
RADIOLOC.
RADIOLOC.
RADIOLOCATION
Mobile
Satellite (S-E)
Mobile
Satellite (S-E)
NOT ALLOCATED
MOBILE
FIXED
MOBILE
SATELLITE
(E-S)
FIXED
SATELLITE
(E-S)
Mobile Satellite (S-E)
FIXED
MET.
FIXED
SATELLITE (S-E) SATELLITE (S-E)
FIXED
FIXED
SATELLITE (S-E)
MOBILE
FIXED
FIXED
FIXED
MOBILE*
608.0
614.0
746
FIXED
FIXED SATELLITE (S-E)
74.0
FIXED
SATELLITE (E-S)
505
510
MARITIME MOBILE
MARITIME
MOBILE
(SHIPS ONLY)
6.525
7.025
7.075
7.125
Fixed
VLF
FIXED
MOBILE
SATELLITE (S-E)
495
MOBILE (DISTRESS AND CALLING)
4.995
5.003
5.005
5.060
AERONAUTICAL MOBILE (R)
FIXED
SPACE RESEARCH (E-S)
FIXED
FIXED
FIXED SATELLITE (S-E)
FIXED
ISM – 6.78 ± .015 MHz
MOBILE
71.0
50.0
AERONAUTICAL MOBILE (OR)
6.70
6.875
3 kHz
FIXED
Aeronautical
Radionavigation
FIXED
FIXED SAT (E-S)
AERONAUTICAL
RADIONAVIGATION
(RADIO BEACONS)
FIXED
FIXED
FIXED
SATELLITE (E-S)
4.85
49.6
MOBILE*
6.425
Aeronautical
Radionavigation
(Radio Beacons)
FIXED
SATELLITE (E-S)
MOBILE
MOBILE
MOBILE
54.0
RADIO ASTRONOMY
6.525
FIXED SATELLITE (S-E)(E-S)
FIXED
FIXED
5.45
5.83
MOBILE
MARITIME
MOBILE
66.0
Aeronautical
Mobile
65.0
RADIORADIO
INTERMOBILE
NAVIGATION SATELLITE NAVIGATION M O B I L E SATELLITE
SATELLITE
MF
INTERSATELLITE
FIXED
FIXED
SATELLITE (E-S)
64.0
435
MOBILE*
BROADCASTING
(TV CHANNELS 2-4)
INTERSATELLITE
FIXED
FIXED
SATELLITE (E-S)
415
4.75
5.65
LAND MOBILE
MARITIME
MOBILE
4.65
4.7
STANDARD FREQ. AND TIME SIGNAL (5000 KHZ)
Space Research
STANDARD FREQ.
5.6
5.85
5.925
335
AERONAUTICAL MOBILE (OR)
46.6
47.0
MOBILE
FIXED
RADIONAVIGATION
4.438
AERONAUTICAL MOBILE (R)
MOBILE
(TV CHANNELS 14 - 20)
3
405
Aeronautical Mobile
AERONAUTICAL
RADIONAVIGATION
FIXED
MOBILE*
43.69
LAND
MOBILE
5.0
5.46
5.47
AERONAUTICAL
RADIONAVIGATION
(RADIO BEACONS)
Aeronautical
Mobile
4.0
4.063
LAND
MOBILE
FIXED
TV BROADCASTING
MOBILE**
Amateur
RADIOLOCATION
RADIOAmateur- sat (s-e) Amateur
LOCATION
Amateur
MOBILE FIXED SAT(E-S)
MARITIME MOBILE
42.0
512.0
5.35
MARITIME
RADIONAVIGATION
(RADIO BEACONS)
54.25
55.78
56.9
57.0
INTERSATELLITE
EARTH
SPACE
EXPLORATION RESEARCH F I X E D M O B I L E * *
SATELLITE
RADIORadioLOCATION
location
Radiolocation
RADIONAVIGATION
MARITIME
Radiolocation
RADIONAVIGATION
MARITIME
METEOROLOGICAL
Radiolocation
RADIONAVIGATION
AIDS
AERONAUTICAL
RADIONAV.
ISM – 5.8 ± .075 GHz
ISM – 61.25 ± .250 GHz
59-64 GHz IS DESIGNATED FOR
UNLICENSED DEVICES
RADIOLOCATION
MOBILE
FIXED SAT (S-E)
Radiolocation
FIXED
40.0
4.99
5.15
5.25
300 kHz
AERO. RADIONAV.
RADIOLOCATION
39.0
FIXED
52.6
59.3
FIXED
FIXED
37.0
37.5
38.0
38.25
LAND MOBILE
BROADCASTING
LAND MOBILE
FIXED
EARTH-ES
INTER- SAT
51.4
Space Research (Passive)
RADIO ASTRONOMY
AERONAUTICAL
RADIONAVIGATION
EARTH
EXPLORATION
SAT. (Passive)
58.2
EARTH
EXPLORATION
SAT. (Passive)
59.0
EARTH
RADIO- INTEREXPLORATION F I X E D M O B I L E SPACE LOC.
RES..
SAT
SAT. (Passive)
MOBILE
MOBILE
460.0
462.5375
462.7375
467.5375
467.7375
470.0
4.94
50.4
SPACE
RESEARCH
(Passive)
FIXED
FIXED
MOBILE**
RES.
EARTH-ES
MOBILE**
INTER
- SAT
FIXED
RADIO ASTRONOMY
MARITIME
MOBILE
SPACE
MOBILE
FIXED
300
325
Maritime
Radionavigation
(Radio Beacons)
AMATEUR
EARTH EXPL-SAT (Passive)
RES.
FIXED
50.2
MOBILE**
SPACE
MOBILE
FIXED
4.8
48.2
HF
INTER- SAT
INTER- SAT
MOBILE
SPACE
RES.
MOBILE
LAND MOBILE
Radio Astronomy
LAND MOBILE
FIXED
47.2
Radiolocation
RES.
MOBILE
SPACE
FIXED
FIXED
FIXED
LAND MOBILE
FIXED
LAND MOBILE
LAND MOBILE FIXED
LAND MOBILE
3 MHz
SPACE
RESEARCH
(Passive)
47.0
36.0
MOBILE
FIXED
402.0
450.0
454.0
455.0
456.0
3.230
3.4
AMATEUR
EARTH
EXPLORATION
SATELLITE
(Passive)
46.9
LAND MOBILE
3.155
FIXED
3.5
401.0
BROADCASTING
(TV CHANNELS 21-36)
MOBILE
4.5
AERONAUTICAL
MOBILE (OR)
MOBILE
4.4
45.5
FIXED
LAND
MOBILE
3.0
3.025
AERONAUTICAL MOBILE (R)
MOBILE*
AERONAUTICAL
MOBILE (R)
LAND
MOBILE
MOBILE
FIXED
400.05
400.15
Amateur
LAND MOBILE
FIXED
LAND MOBILE
LAND MOBILE
FIXED
LAND MOBILE
MOBILE
FIXED
43.5
MOBILE
FIXED
SPACE RESEARCH
VHF
FX
SAT(E-S)
EARTH
EXPLORATION
SATELLITE
FI XED
MOBILE
SATELLITE (E-S) SATELLITE (E-S)
Met-Satellite Earth Expl Sat
(E-S)
(E-S)
AERONAUTICAL
RADIONAVIGATION
LAND
MOBILE
FIXED
ISM – 40.68 ± .02 MHz
FX
SAT(E-S)
MOBILE
Earth
Expl.
Met-Satellite
Earth Expl Sat
(E-S)Satellite(E-S)
(E-S)
MET-SAT. EARTH EXPL
SAT. (E-S)
(E-S)
Meteorological
Satellite (S-E)
AMATEUR SATELLITE
MOBILE
FIXED
MET-SAT. EARTH EXPL
(E-S)
SAT. (E-S)
MET. AIDS
(Radiosonde)
RADIOLOCATION
FIXED
AMATEUR
FIXED
(S-E)
MET. AIDS SPACE OPN.
(Radio(S-E)
sonde)
420.0
MOBILE
MOBILE
MOBILE
SAT (E-S).
MOB. SAT(E-S) MOBILE
FIXED
MET. SAT.
(S-E)
Space Opn.
(S-E)
SPACE RES.
30.56
35.0
403.0
4.2
FIXED
SATELLITE (S-E)
MOBILE
SATELLITE (E-S)
FIXED
SATELLITE (E-S)
MOBILE.
SAT. (S-E)
30.0
FIXED
399.9
METEOROLOGICAL AIDS (RADIOSONDE)
406.0
MOBILE SATELLITE (E-S)
406.1
RADIO
FIXED
MOBILE
ASTRONOMY
410.0
SPACE RESEARCH
FIXED
MOBILE
(S-S)
40.0
R AD IONAV.
SATELLITE
FIXED
MET. AIDS
(Radiosonde)
3.7
MOBILE
ASTRONOMY
30 MHz
39.5
EARTH
Earth
SPACE
F I X E D MOBILE
EXPL
Expl.
SAT
RES. (E-S) Sat (s - e) SAT (E-S)
SAT.
40.5
BROAD- FX-SAT Fixed Mobile
BCST
C A S T I N G (S-E)
SAT.
41.0
BROADBCST
MOBILE
FIXED
CASTING
SAT.
42.5
RADIO
FIXED
F I X E D M O B I L E * * SATELLITE
(E-S)
RADIONAV.SAT.
STD. FREQ. & TIME SIGNAL SAT. (400.1 MHz)
3.65
FIXED
MOBILE
33.0
MOBILE
FIXED
38.0
LAND
MOBILE
34.0
MOBILE SATELLITE (E-S)
RADIONAVIGATION SATELLITE
37.6
38.6
MOBILE
SATELLITE
FIXED
MOBILE
SATELLITE S A T .
MOBILE
FIXED SAT.
(S-E)
MOBILE**
3.6
FIXED
SATELLITE
(S-E)
FIXED
FIXED-SATELLITE
Radiolocation
MOBILE
37.0
MOBILE
SATELLITE
36.0
FIXED SAT.
(S-E)
FIXED
FIXED
SAT. (S-E)
FIXED
FIXED
SATELLITE (S-E)
RADIOLOCATION
AERO. RADIONAV.(Ground)
MOBILE
FIXED
FIXED
328.6
335.4
Radiolocation
RADIOLOCATION
FIXED
32.0
MOBILE
AERONAUTICAL RADIONAVIGATION
3.5
AERONAUTICAL
RADIONAVIGATION
(Ground)
SPACE RESEARCH
(space-to-Earth)
SPACE
RES.
FIXED
322.0
Amateur
SPACE RE. EARTH EXPL.
.(Passive) SAT. (Passive)
MOBILE
Radiolocation
33.0
33.4
Radiolocation
MOBILE
MOBILE
RADIOLOCATION
FIXED
300.0
FIXED
3.3
32.0
RADIOLOCATION
FIXED
31.8
INTER-SATELLITE
RADIONAVIGATION
F I X E D MOBILE
3.1
RADIONAVIGATION 32.3
INTER- SAT
RADIONAVIGATION
Radiolocation
FIXED
RADIONAVIGATION
3.0
MARITIME
RADIONAVIGATION
300 MHz
SPACE
RESEARCH (deep space)
SPACE RES.
31.3
RADIOLOCATION
MOBILE
EARTH
EXPLORATION
SAT. (Passive)
Radiolocation
31.0
FIXED
SPACE
RESEARCH
(Passive)
3 GHz
MOBILE
SATELLITE
(E-S)
FIXED
SATELLITE
(E-S)
RADIO
ASTRONOMY
UHF
SHF
EHF
30 GHz
Figure 1-17 Individual bands of the radio spectrum and their primary allocations in the U.S. [See expandable version on book website:
em8e.eecs.umich.edu.]
30.0
Standard
Frequency and
Time Signal
Satellite (S-E)
Stand. Frequency
and Time Signal
Satellite (S-E)
1-6 REVIEW OF COMPLEX NUMBERS
33
Module 1.3 Phase Lead/Lag Examine sinusoidal waveforms with different values of the reference phase constant φ0 .
We now highlight some of the properties of complex algebra
that will be encountered in future chapters.
(z)
Equality: If two complex numbers z1 and z2 are given by
z1 = x1 + jy1 = |z1 |e jθ1 ,
z2 = x2 + jy2 = |z2 |e jθ2 ,
(1.44)
θ = tan−1 (y/x)
|z|
Addition:
(1.46)
θ
x
(z)
Figure 1-18 Relation between rectangular and polar representations of a complex number z = x + jy = |z|e jθ .
Multiplication:
z1 z2 = (x1 + jy1 )(x2 + jy2 )
= (x1 x2 − y1y2 ) + j(x1 y2 + x2 y1 ),
|z| = + x2 + y2
(1.45)
then z1 = z2 if and only if x1 = x2 and y1 = y2 or, equivalently,
|z1 | = |z2 | and θ1 = θ2 .
z1 + z2 = (x1 + x2 ) + j(y1 + y2).
z
y
x = |z| cos θ
y = |z| sin θ
(1.47a)
34
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
or
∗
(a) express V and
√ I in polar form, and find (b) V I, (c) V I ,
(d) V /I, and (e) I .
z1 z2 = |z1 |e jθ1 · |z2 |e jθ2
= |z1 ||z2 |e j(θ1 +θ2 )
= |z1 ||z2 |[cos(θ1 + θ2 ) + j sin(θ1 + θ2 )].
(1.47b)
θ V = tan−1 (−4/3) = −53.1◦,
Division: For z2 6= 0,
◦
z1
x1 + jy1 (x1 + jy1 ) (x2 − jy2 )
=
=
·
z2
x2 + jy2 (x2 + jy2 ) (x2 − jy2 )
=
(x1 x2 + y1 y2 ) + j(x2 y1 − x1y2 )
,
x22 + y22
(1.48a)
or
V = |V |e jθV = 5e− j53.1 = 5 −53.1◦ ,
p
√
+
|I| = 22 + 32 = + 13 = 3.61.
Since I = (−2 − j3) is in the third quadrant in the complex
plane (Fig. 1-19),
θI = 180◦ + tan−1 32 = 236.3◦,
I = 3.61 236.3◦ .
|z1 |e jθ1
|z1 | j(θ1 −θ2 )
z1
=
=
e
j
θ
2
z2
|z2 |e
|z2 |
=
Solution:
√
+
(a) |V | = p
VV ∗
√
= + (3 − j4)(3 + j4) = + 9 + 16 = 5,
◦
|z1 |
[cos(θ1 − θ2 ) + j sin(θ1 − θ2 )]. (1.48b)
|z2 |
(b) V I = 5e− j53.1 × 3.61e j236.3
= 18.03e j(236.3
◦−53.1◦ )
◦
◦
= 18.03e j183.2 .
◦
(c) V I ∗ = 5e− j53.1 × 3.61e− j236.3
◦
◦
◦
= 18.03e− j289.4 = 18.03e j70.6 .
Powers: For any positive integer n,
◦
zn = (|z|e jθ )n
◦
5e− j53.1
V
− j289.4◦
= 1.39e j70.6 .
=
◦ = 1.39e
j236.3
I
3.61e
√
√
◦
(e) I = 3.61e j236.3
√
◦
◦
= ± 3.61 e j236.3 /2 = ±1.90e j118.15 .
(d)
= |z|n e jnθ = |z|n (cos nθ + j sin nθ ),
(1.49)
z1/2 = ±|z|1/2 e jθ /2 = ±|z|1/2 [cos(θ /2) + j sin(θ /2)].
(1.50)
Useful Relations:
−1 = e jπ = e− jπ = 1 180◦ ,
j = e jπ /2 = 1 90◦ ,
(1.51)
− j = −e jπ /2 = e− jπ /2 = 1 −90◦ ,
p
±(1 + j)
,
j = (e jπ /2 )1/2 = ±e jπ /4 = √
2
p
±(1 − j)
.
− j = ±e− jπ /4 = √
2
(1.52)
(1.53)
(1.54)
−2
θI
3
θV
|I |
Example 1-3:
Working with Complex
Numbers
−3
I
−4
Given two complex numbers
V
Figure 1-19 Complex numbers V and I in the complex plane
(Example 1-3).
V = 3 − j4,
|V |
I = −(2 + j3),
1-6 REVIEW OF COMPLEX NUMBERS
Technology Brief 2: Solar Cells
A solar cell is a photovoltaic device that converts solar
energy into electricity. The conversion process relies
on the photovoltaic effect, which was first reported
by 19-year-old Edmund Bequerel in 1839 when he
observed that a platinum electrode produced a small
current if exposed to light. The photovoltaic effect is
often confused with the photoelectric effect; they are
interrelated, but not identical (Fig. TF2-1).
The photoelectric effect explains the mechanism
responsible for why an electron is ejected by a material
in consequence to a photon incident upon its surface
(Fig. TF2-1(a)). For this to happen, the photon energy E
(which is governed by its wavelength through E = hc/λ ,
with h being Planck’s constant and c the velocity of light)
has to exceed the binding energy with which the electron
is held by the material. For his 1905 quantum-mechanical
model of the photoelectric effect, Albert Einstein was
awarded the 1921 Nobel Prize in physics.
Whereas a single material is sufficient for the photoelectric effect to occur, at least two adjoining materials
with different electronic properties (to form a junction
e
Photon
_
Metal
(a) Photoelectric effect
Photon
e
_
n-type
p-type
Load
I
(b) Photovoltaic effect
Figure TF2-1 Comparison of photoelectric effect with the
photovoltaic effect.
35
that can support a voltage across it) are needed to
establish a photovoltaic current through an external
load (Fig. TF2-1(b)). Thus, the two effects are governed
by the same quantum-mechanical rules associated with
how photon energy can be used to liberate electrons
away from their hosts, but the followup step of what
happens to the liberated electrons is different in the two
cases.
The PV Cell
Today’s photovoltaic (PV) cells are made of semiconductor materials. The basic structure of a PV cell consists of
a p-n junction connected to a load (Fig. TF2-2).
Typically, the n-type layer is made of silicon doped
with a material that creates an abundance of negatively
charged atoms, and the p-type layer also is made of
silicon but doped with a different material that creates an
abundance of holes (atoms with missing electrons). The
combination of the two layers induces an electric field
across the junction, so when an incident photon liberates
an electron, the electron is swept under the influence
of the electric field through the n-layer and out to the
external circuit connected to the load.
The conversion efficiency of a PV cell depends on
several factors, including the fraction of the incident light
that gets absorbed by the semiconductor material, as
opposed to getting reflected by the n-type front surface
or transmitted through to the back conducting electrode.
To minimize the reflected component, an antireflective
coating usually is inserted between the upper glass cover
and the n-type layer (Fig. TF2-2).
The PV cell shown in Fig. TF2-2 is called a singlejunction cell because it contains only one p-n junction.
The semiconductor material is characterized by a quantity called its band gap energy, which is the amount of
energy needed to free an electron away from its host
atom. For that to occur, the wavelength of the incident
photon (which in turn defines its energy) has to be such
that the photon’s energy exceeds the band gap of the
material. Solar energy extends over a broad spectrum,
so only a fraction of the solar spectrum (photons with
energies greater than the band gap) is absorbed by
a single-junction material. To overcome this limitation,
multiple p-n layers can be cascaded together to form a
multijunction PV device (Fig. TF2-3). The cells usually
are arranged such that the top cell has the highest band
gap energy, thereby capturing the high-energy (shortwavelength) photons, followed by the cell with the next
lower band gap, and so on.
36
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Antireflective coating
Glass cover
Light photons
Front conducting electrode
−
n-type layer
(silicon)
+
p-n junction
p-type layer (silicon)
Back conducting electrode
Figure TF2-2 Basic structure of a photovoltaic cell.
IR
400
500
600
Wavelength (nm)
700
800
InGaP
InGaAs
Ge
Figure TF2-3 In a multijunction PV device, different layers absorb different parts of the light spectrum.
1-6 REVIEW OF COMPLEX NUMBERS
PV cell
37
PV array
PV module
Figure TF2-4 PV cells, modules, and arrays.
◮ The multijunction technique offers an improvement
in conversion efficiency of 2–4 times over that of the
single-junction cell. However, the fabrication cost is
significantly greater. ◭
PV array
dc
Modules, Arrays, and Systems
A photovoltaic module consists of multiple PV cells
connected together in order to supply electrical power
at a specified voltage level, such as 12 or 24 V. The
combination of multiple modules generates a PV array
(Fig. TF2-4). The amount of generated power depends
on the intensity of the intercepted sunlight, the total area
of the module or array, and the conversion efficiencies
of the individual cells. If the PV energy source is to
serve multiple functions, it is integrated into an energy
management system that includes a dc to ac current
converter and batteries to store energy for later use
(Fig. TF2-5).
Battery
storage
system
dc
dc to ac
dc/ac inverter
ac
Figure TF2-5
system.
Components of a large-scale photovoltaic
38
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Exercise 1-9: Express the following complex functions in
R
polar form:
+
z1 = (4 − j3)2,
z2 = (4 − j3)1/2.
√
Answer: z1 = 25 −73.7◦ , z2 = ± 5 −18.4◦ . (See
C
υs(t)
i
−
EM
.)
Figure 1-20 RC circuit connected to a voltage source υs (t).
√
Exercise 1-10: Show that 2 j = ±(1 + j).
Answer: (See
EM
.)
gives the following loop equation:
R i(t) +
Phasor analysis is a useful mathematical tool for solving
problems involving linear systems in which the excitation
is a periodic time function. Many engineering problems are
cast in the form of linear integro-differential equations. If the
excitation, more commonly known as the forcing function,
varies sinusoidally with time, the use of phasor notation to
represent time-dependent variables allows us to convert a
linear integro-differential equation into a linear equation with
no sinusoidal functions, thereby simplifying the method of
solution. After solving for the desired variable, such as the
voltage or current in a circuit, conversion from the phasor
domain back to the time domain provides the desired result.
The phasor technique also can be used to analyze linear systems when the forcing function is a (nonsinusoidal)
periodic time function, such as a square wave or a sequence
of pulses. By expanding the forcing function into a Fourier
series of sinusoidal components, we can solve for the desired
variable using phasor analysis for each Fourier component of
the forcing function separately. According to the principle of
superposition, the sum of the solutions due to all of the Fourier
components gives the same result as one would obtain had the
problem been solved entirely in the time domain without the
aid of Fourier representation. The obvious advantage of the
phasor–Fourier approach is simplicity. Moreover, in the case
of nonperiodic source functions, such as a single pulse, the
functions can be expressed as Fourier integrals, and a similar
application of the principle of superposition can be used as
well.
The simple RC circuit shown in Fig. 1-20 contains a
sinusoidally time-varying voltage source given by
υs (t) = V0 sin(ω t + φ0 ),
(1.55)
where V0 is the amplitude, ω is the angular frequency, and φ0
is a reference phase. Application of Kirchhoff’s voltage law
Z
1
i(t) dt = υs (t).
C
(time domain)
1-7 Review of Phasors
(1.56)
Our objective is to obtain an expression for the current i(t). We
can do this by solving Eq. (1.56) in the time domain, which is
somewhat cumbersome because the forcing function υs (t) is a
sinusoid. Alternatively, we can take advantage of the phasordomain solution technique as follows.
1-7.1
Solution Procedure
Step 1: Adopt a cosine reference
To establish a phase reference for all time-varying currents and
voltages in the circuit, the forcing function is expressed as a
cosine (if not already in that form). In the present example,
υs (t) = V0 sin(ω t + φ0 )
π
− ω t − φ0
= V0 cos
2
π
= V0 cos ω t + φ0 −
,
2
(1.57)
where we used the properties sin x = cos(π /2 − x) and
cos(−x) = cos x.
Step 2: Express time-dependent variables as phasors
Any cosinusoidally time-varying function z(t) can be
expressed as
h
i
(1.58)
z(t) = Re Ze e jω t ,
where Ze is a time-independent function called the phasor of
the instantaneous function z(t). To distinguish instantaneous
quantities from their phasor counterparts, a tilde (∼) is added
1-7 REVIEW OF PHASORS
39
over the letter representing a phasor. For the voltage υs (t)
given by Eq. (1.57), its phasor equivalent Ves is given by
Ves = V0 e j(φ0 −π /2) .
(1.59)
es has the same amplitude as υs (t), and its exponential
Thus, V
function is j(φ0 − π /2), where (φ0 − π /2) is the phase part
of the cosine function in Eq. (1.57) without the ω t term. To
verify that the expression for Ves given by Eq. (1.59) is indeed
the phasor equivalent of υs (t), we apply the right-hand side of
Eq. (1.58):
Re[Ves e jω t ] = Re[V0 e j(φ0 −π /2) · e jω t ]
= V0 cos(ω t + φ0 − π /2),
which is, indeed, the expression for υs (t) given by Eq. (1.57).
The time-domain and phasor-domain correspondence can be
expressed in the general form
Ves = V0 e jφ ,
υs (t) = V0 cos(ω t + φ )
(1.60)
for any phase angle φ . Hence, φ may be a constant, such
as φ0 − π /2, or it may be a function of one or more spatial
variables, such as φ = β x.
The phasor Ves , corresponding to the time function υs (t),
contains amplitude and phase information but it is independent
of the time variable t. Next we define the unknown variable i(t)
in terms of a phasor equivalent Ie as
e jω t ).
i(t) = Re(Ie
(1.61)
If the equation we are trying to solve contains derivatives or
integrals, we use the following two properties:
i
di
d h
e jω t ) = Re d (Ie
e jω t ) = Re[ jω Ie
e jω t ],
Re(Ie
=
dt
dt
dt
(1.62)
and
Z
i dt =
Z
e jω t ) dt
Re(Ie
= Re
Z
e jω t dt
Ie
= Re
!
Ie jω t
e
.
jω
(1.63)
◮ Thus, differentiation of the time function i(t) with
respect to time is equivalent to multiplication of its phasor
Ieby jω , and integration is equivalent to division by jω . ◭
Step 3: Recast the differential / integral equation in phasor
form
Upon using the form of Eq. (1.58) to represent i(t) and υs (t)
in Eq. (1.56) in terms of their phasor equivalents, we have
!
e
1
I
e jω t ) + Re
(1.64)
R Re(Ie
e jω t = Re(Ves e jω t ).
C
jω
Combining all three terms under the same real-part (Re)
operator leads to
1
jω t
e
e
Re
R+
= 0.
(1.65a)
I − Vs e
jω C
Had we adopted a sine reference—instead of a cosine
reference—to define sinusoidal functions, the preceding treatment would have led to the result
1
jω t
e
e
= 0.
(1.65b)
Im
R+
I − Vs e
jω C
Since both the real and imaginary parts of the expression inside
the curly brackets are zero, the expression itself must be zero.
Moreover, since e jω t 6= 0, it follows that
1
es
e
=V
I R+
jω C
(phasor domain).
(1.66)
The time factor e jω t has disappeared because it was contained
in all three terms. Equation (1.66) is the phasor-domain equivalent of Eq. (1.56).
Step 4: Solve the phasor-domain equation
From Eq. (1.66), the phasor current Ie is given by
Ie =
es
V
.
R + 1/( jω C)
(1.67)
Before we apply the next step, we need to convert the righthand side of Eq. (1.67) into the form I0 e jθ with I0 being a real
quantity. Thus,
jω C
j(φ0 −π /2)
e
I = V0 e
1 + jω RC
"
#
ω Ce jπ /2
j(φ0 −π /2)
p
= V0 e
1 + ω 2R2C2 e jφ1
V0 ω C
=p
e j(φ0 −φ1 ) ,
1 + ω 2 R2C 2
(1.68)
where we have used the identity j = e jπ /2 . The phase angle
associated with (1 + jω RC) is φ1 = tan−1 (ω RC) and lies in
the first quadrant of the complex plane.
40
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Phasor Technique Solution Procedure
(1) Adopt a cosine reference for all sinusoidally varying
quantities, as in
Table 1-5 Time-domain sinusoidal functions z(t) and
e where
their cosine-reference phasor-domain counterparts Z,
e jω t ].
z(t) = Re[Ze
Ze
z(t)
x(t) = A cos(ω t + φ0 ).
• Coefficient A should be positive. If not, retain
its positive magnitude and add or subtract π
from φ0 .
• If the time function is a sine instead of a cosine,
convert it using Eq. (1.57).
A cos ω t
A cos(ω t + φ0 )
A cos(ω t + β x + φ0 )
Ae−α x cos(ω t + β x + φ0 )
A sin ω t
A sin(ω t + φ0 )
A
Ae jφ0
Ae j(β x+φ0 )
Ae−α x e j(β x+φ0 )
Ae− jπ /2
Ae j(φ0 −π /2)
d
(z(t))
dt
jω Ze
(2) Convert time-dependent quantities into phasor equivalents:
Ze
z(t)
d
[A cos(ω t + φ0 )]
dt
(Table 1-5).
(3) Cast equations in the phasor domain, and then solve
them for the quantity of interest.
(4) If necessary, modify the solution Ye into the form
Ye = Be jθ ,
Z
z(t) dt
Z
A sin(ω t + φ0 ) dt
jω Ae jφ0
1 e
Z
jω
1
Ae j(φ0 −π /2)
jω
where B is a positive, real quantity.
(5) Convert the phasor-domain solution to the time
domain:
y(t) = Re[Be jθ · e jω t ] = B cos(ω t + θ )
Example 1-4: RL Circuit
The voltage source of the circuit shown in Fig. 1-21 is given
by
(V).
(1.70)
υs (t) = 5 sin(4 × 104t − 30◦)
Step 5: Convert back to the time domain
Obtain an expression for the voltage across the inductor.
To find i(t), we simply apply Eq. (1.61). That is, we multiply
the phasor Ie given by Eq. (1.68) by e jω t and then take the real
part:
"
e jω t ] = Re p V0 ω C
e j(φ0 −φ1 ) e jω t
i(t) = Re[Ie
2
2
2
1+ω R C
#
V0 ω C
=p
cos(ω t + φ0 − φ1 ). (1.69)
1 + ω 2 R2C 2
In summary, we converted all time-varying quantities into
the phasor domain, solved for the phasor Ie of the desired
instantaneous current i(t), and then converted back to the time
domain to obtain an expression for i(t). Table 1-5 provides
a summary of some time-domain functions and their phasordomain equivalents.
i
R=6Ω
+
+
υs(t)
υL
L = 0.2 mH
−
−
Figure 1-21 RL circuit (Example 1-4).
Solution: The voltage loop equation of the RL circuit is
Ri + L
di
= υs (t).
dt
(1.71)
1-7 REVIEW OF PHASORS
41
Before converting Eq. (1.71) into the phasor domain, we
express Eq. (1.70) in terms of a cosine reference:
υs (t) = 5 sin(4 × 104t − 30◦)
= 5 cos(4 × 104t − 120◦)
(V).
(1.72)
The coefficient of t specifies the angular frequency as
ω = 4 × 104 (rad/s). Per the second entry in Table 1-5, the
voltage phasor corresponding to υs (t) is
Ves = 5e
− j120◦
(1.73)
e we have
Solving for the current phasor I,
Find υ (t).
Answer:
◦
◦
(See
◦
5e− j120
5e− j120
− j173.1◦
=
◦ = 0.5e
j53.1
6 + j8
10e
(A).
The voltage phasor across the inductor is related to Ie by
4
= j4 × 10 × 2 × 10
= 4e
(a) Ie = 150/(R + jω L) = 0.3 −36.9◦ (A),
(b) i(t) = 0.3 cos(ω t − 36.9◦) (A). (See EM .)
e = j5 V.
Exercise 1-12: A phasor voltage is given by V
5e− j120
=
6 + j4 × 104 × 2 × 10−4
VeL = jω LIe
Exercise 1-11: A series RL circuit is connected to a
Answer:
es .
RIe+ jω LIe = V
=
How is the phasor technique
used when the forcing function is a nonsinusoidal periodic
waveform, such as a train of pulses?
Concept Question 1-13:
voltage source given by υs (t) = 150 cos ω t (V). Find
(a) the phasor current Ie and (b) the instantaneous current
i(t) for R = 400 Ω, L = 3 mH, and ω = 105 rad/s.
(V),
and the phasor equation corresponding to Eq. (1.71) is
es
V
Ie =
R + jω L
Why is the phasor technique
useful? When is it used? Describe the process.
Concept Question 1-12:
j(90◦ −173.1◦ )
−4
= 4e
× 0.5e
− j83.1◦
− j173.1◦
.)
υ (t) = 5 cos(ω t + π /2) = −5 sin ω t
(V).
1-7.2 Traveling Waves in the Phasor Domain
According to Table 1-5, if we set φ0 = 0, its third entry
becomes
A cos(ω t + β x)
Ae jβ x .
(1.74)
From the discussion associated with Eq. (1.31), we concluded
that A cos(ω t + β x) describes a wave traveling in the negative
x direction.
(V),
and the corresponding instantaneous voltage υL (t) is therefore
i
h
υL (t) = Re VeL e jω t
h
i
◦
4
= Re 4e− j83.1 e j4×10 t
= 4 cos(4 × 104t − 83.1◦)
EM
(V).
◮ In the phasor domain, a wave of amplitude A traveling
in the positive x direction in a lossless medium with a
phase constant β is given by the negative exponential
Ae− jβ x . Conversely, a wave traveling in the negative x
direction is given by Ae jβ x . Thus, the sign of x in the
exponential is opposite to the direction of travel. ◭
42
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
Chapter 1 Summary
Concepts
• Electromagnetics consists of three branches:
• Electromagnetics is the study of electric and magnetic
phenomena and their engineering applications.
• The International System of Units consists of the six
fundamental dimensions listed in Table 1-1. The units
of all other physical quantities can be expressed in
terms of the six fundamental units.
• The four fundamental forces of nature are nuclear,
weak-interaction, electromagnetic, and gravitational
forces.
• The source of the electric field quantities E and D
is the electric charge q. In a material, E and D
are related by D = ε E, where ε is the electrical permittivity of the material. In free space,
ε = ε0 ≈ (1/36π ) × 10−9 (F/m).
• The source of the magnetic field quantities B and H is
the electric current I. In a material, B and H are related
by B = µ H, where µ is the magnetic permeability of
the medium. In free space, µ = µ0 = 4π × 10−7 (H/m).
(1) electrostatics, which pertains to stationary charges,
(2) magnetostatics, which pertains to dc currents, and
(3) electrodynamics, which pertains to time-varying
currents.
• A traveling wave is characterized by a spatial wavelength λ , a time period T , and a phase velocity
up = λ /T .
• An electromagnetic (EM) wave consists of oscillating
electric and magnetic field intensities and
√ travels in free
space at the velocity of light: c = 1/ ε0 µ0 . The EM
spectrum encompasses gamma rays, X-rays, visible
light, infrared waves, and radio waves.
• Phasor analysis is a useful mathematical tool for solving problems involving time-periodic sources.
Mathematical and Physical Models
Electric field due to charge q in free space
q
E = R̂
4πε0 R2
Complex numbers
•
Magnetic field due to current I in free space
µ0 I
B = φ̂φ
2π r
Plane wave y(x,t) = Ae−α x cos(ω t − β x + φ0)
• α = 0 in lossless medium
ω
• phase velocity up = f λ = β
• ω = 2π f ; β = 2π /λ
• φ0 = phase reference
e jθ = cos θ + j sin θ
• Rectangular-polar relations
y = |z| sin θ ,
x = |z| cos θ ,
p
+ 2
2
|z| = x + y ,
θ = tan−1 (y/x)
Phasor-domain equivalents
PROBLEMS
Section 1-4: Traveling Waves
∗ 1.1 Aharmonic wave traveling along a string is generated by
an oscillator that completes 180 vibrations per minute. If it is
observed that a given crest, or maximum, travels 300 cm in
10 s, what is the wavelength?
1.2 For the pressure wave described in Example 1-1, plot the
∗ Answer(s) available in Appendix E.
Euler’s identity
See Table 1-5
following:
(a) p(x,t) versus x at t = 0
(b) p(x,t) versus t at x = 0
Be sure to use appropriate scales for x and t so that each of
your plots covers at least two cycles.
∗ 1.3 A 2 kHz sound wave traveling in the x direction in air
was observed to have a differential pressure p(x,t) = 10 N/m2
at x = 0 and t = 50 µ s. If the reference phase of p(x,t) is 36◦ ,
find a complete expression for p(x,t). The velocity of sound in
air is 330 m/s.
PROBLEMS
43
Important Terms
Provide definitions or explain the meaning of the following terms:
angular velocity ω
attenuation constant α
attenuation factor
Biot–Savart law
complex conjugate
complex number
conductivity σ
constitutive parameters
continuous periodic wave
Coulomb’s law
dielectric constant
dynamic
electric dipole
electric field intensity E
electric flux density D
electric polarization
electrical force
electrical permittivity ε
electrodynamics
electrostatics
EM spectrum
Euler’s identity
forcing function
fundamental dimensions
instantaneous function
law of conservation of electric charge
LCD
liquid crystal
lossless or lossy medium
magnetic field intensity H
magnetic flux density B
magnetic force
magnetic permeability µ
magnetostatics
microwave band
monochromatic
nonmagnetic materials
perfect conductor
perfect dielectric
periodic
phase
phase constant (wave number) β
phase lag and lead
phase velocity (propagation
velocity) up
phasor
plane wave
principle of linear superposition
reference phase φ0
relative permittivity or
dielectric constant εr
SI system of units
static
transient wave
velocity of light c
wave amplitude
wave frequency f
wave period T
waveform
wavelength λ
1.4 A wave traveling along a string is given by
y(x,t) = 2 sin(4π t + 10π x)
y
(cm),
where x is the distance along the string in meters and y is
the vertical displacement. Determine: (a) the direction of wave
travel, (b) the reference phase φ0 , (c) the frequency, (d) the
wavelength, and (e) the phase velocity.
Incident wave
x
1.5 Two waves, y1 (t) and y2 (t), have identical amplitudes
and oscillate at the same frequency, but y2 (t) leads y1 (t) by
a phase angle of 60◦ . If
Figure P1.7 Wave on a string tied to a wall at x = 0
y1 (t) = 4 cos(2π × 103t),
(Problem 1.7).
write the expression appropriate for y2 (t) and plot both functions over the time span from 0 to 2 ms.
∗ 1.6 The height of an ocean wave is described by the function
y(x,t) = 1.5 sin(0.5t − 0.6x)
x=0
When wave y1 (x,t) arrives at the wall, a reflected wave y2 (x,t)
is generated. Hence, at any location on the string, the vertical
displacement ys is the sum of the incident and reflected waves:
(m).
Determine the phase velocity and wavelength, and then sketch
y(x,t) at t = 2s over the range from x = 0 to x = 2λ .
1.7 A wave traveling along a string in the +x direction is
given by
y1 (x,t) = A cos(ω t − β x),
where x = 0 is the end of the string, which is tied rigidly to a
wall, as shown in Fig. P1.7.
ys (x,t) = y1 (x,t) + y2(x,t).
(a) Write an expression for y2 (x,t), keeping in mind its
direction of travel and the fact that the end of the string
cannot move.
(b) Generate plots of y1 (x,t), y2 (x,t), and ys (x,t) versus x
over the range −2λ ≤ x ≤ 0 at ω t = π /4 and at ω t = π /2.
44
CHAPTER 1 INTRODUCTION: WAVES AND PHASORS
1.8 Two waves on a string are given by the functions:
y1 (x,t) = 4 cos(20t − 30x)
y2 (x,t) = −4 cos(20t + 30x)
(cm)
(a) Find the frequency, wavelength, and phase velocity of the
wave.
(b) At z = 2 m, the amplitude of the wave was measured to
be 2 V. Find α .
(cm),
where x is in centimeters. The waves are said to interfere
constructively when their superposition |ys | = |y1 + y2 | is a
maximum, and they interfere destructively when |ys | is a
minimum.
∗ (a) What are the directions of propagation of waves y (x,t)
1
and y2 (x,t)?
(b) At t = (π /50) s, what location x do the two waves
interfere constructively, and what is the corresponding
value of |ys |?
(c) At t = (π /50) s, what location x do the two waves
interfere destructively, and what is the corresponding
value of |ys |?
1.9 Find expressions for y(x,t) for a sinusoidal wave traveling along a string in the negative x direction, given that
ymax = 40 cm, λ = 30 cm, f = 10 Hz, and
(a) y(x, 0) = 0 at x = 0
(b) y(x, 0) = 0 at x = 3.75 cm
∗ 1.10 An oscillator that generates a sinusoidal wave on a
string completes 20 vibrations in 50 s. The wave peak is
observed to travel a distance of 2.8 m along the string in 5 s.
What is the wavelength?
∗ 1.14 A certain electromagnetic wave traveling in seawater
was observed to have an amplitude of 98.02 (V/m) at a depth
of 10 m, and an amplitude of 81.87 (V/m) at a depth of 100 m.
What is the attenuation constant of seawater?
1.15 A laser beam traveling through fog was observed to
have an intensity of 1 (µ W/m2 ) at a distance of 2 m from
the laser gun and an intensity of 0.2 (µ W/m2 ) at a distance
of 3 m. Given that the intensity of an electromagnetic wave is
proportional to the square of its electric-field amplitude, find
the attenuation constant α of fog.
1.16 In Module 1.2, set f = 1 Hz, λ = 2 cm, and α = 0.5
Np/cm. Estimate the amplitude y(x) of the blue wave at its first
maximum (at x = 0) and at its second maximum (at x = 2 cm).
Compute the ratio y(2 cm)/y(0). Does the result match your
expectations given the values of f , λ , and α ?
Section 1-5: Complex Numbers
1.17 Complex numbers z1 and z2 are given as
z1 = 3 − j2,
z2 = −4 + j3.
1.11 The vertical displacement of a string is given by the
harmonic function:
y(x,t) = 2 cos(16π t − 20π x)
(a) Express z1 and z2 in polar form.
(m),
where x is the horizontal distance along the string in meters.
Suppose a tiny particle were attached to the string at x = 5 cm.
Obtain an expression for the vertical velocity of the particle as
a function of time.
∗ 1.12 Given two waves characterized by
∗ (c) Determine the product z z in polar form.
1 2
(d) Determine the ratio z1 /z2 in polar form.
(e) Determine z31 in polar form.
1.18 Evaluate each of the following complex numbers and
express the result in rectangular form:
y1 (t) = 3 cos ω t,
y2 (t) = 3 sin(ω t + 60◦),
does y2 (t) lead or lag y1 (t) and by what phase angle?
1.13 The voltage of an electromagnetic wave traveling on a
transmission line is given by
υ (z,t) = 5e−α z sin(4π × 109t − 20π z)
(b) Find |z1 | by first applying Eq. (1.41) and then by applying
Eq. (1.43).
(V),
where z is the distance in meters from the generator.
(a) z1 = 8e jπ /3
∗ (b) z = √3 e j3π /4
2
(c) z3 = 2e− jπ /2
(d) z4 = j3
(e) z5 = j−4
(f) z6 = (1 − j)3
(g) z7 = (1 − j)1/2
PROBLEMS
45
1.19 Complex numbers z1 and z2 are given by
z1 = −3 + j2,
z2 = 1 − j2.
Determine (a) z1 z2 , (b) z1 /z∗2 , (c) z21 , and (d) z1 z∗1 all in polar
form.
1.20 If z = −2 + j4, determine the following quantities in
polar form:
(a) 1/z
(b) z3
∗ (c) |z|2
(d) Im{z}
(e) Im{z∗ }
1.21 Find complex numbers t = z1 + z2 and s = z1 − z2 , both
in polar form, for each of the following pairs:
(a) z1 = 2 + j3 and z2 = 1 − j2
(b) z1 = 3 and z2 = − j3
(c) z1 = 3 30◦ and z2 = 3 −30◦
∗ (d) z = 3 30◦ and z = 3 −150◦
1
2
1.22 Complex numbers z1 and z2 are given by
z1 = 5 −60◦ ,
(c) Ie = (6 + j8) (A)
∗ (d) Ie = −3 + j2 (A)
(e) Ie = j (A)
(f) Ie = 2e jπ /6 (A)
1.28 Find the phasors of the following time functions:
(a) υ (t) = 9 cos(ω t − π /3) (V)
(b) υ (t) = 12 sin(ω t + π /4) (V)
(c) i(x,t) = 5e−3x sin(ω t + π /6) (A)
∗ (d) i(t) = −2 cos(ω t + 3π /4) (A)
(e) i(t) = 4 sin(ω t + π /3) + 3 cos(ω t − π /6) (A)
1.29 A series RLC circuit is connected to a generator with a
voltage υs (t) = V0 cos(ω t + π /3) (V).
(a) Write the voltage loop equation in terms of the current
i(t), R, L, C, and υs (t).
(b) Obtain the corresponding phasor-domain equation.
(c) Solve the equation to obtain an expression for the phasor
e
current I.
1.30 The voltage source of the circuit shown in Fig. P1.30 is
given by
z2 = 4 45◦ .
(a)
(b)
(c)
(d)
(e)
1.27 Find the instantaneous time sinusoidal functions corresponding to the following phasors:
e = −5e jπ /3 (V)
(a) V
e = j6e− jπ /4 (V)
(b) V
υs (t) = 25 cos(4 × 104t − 45◦)
Determine the product z1 z2 in polar form.
Determine the product z1 z∗2 in polar form.
Determine the ratio z1 /z2 in polar form.
Determine the ratio z∗1 /z∗2 in polar form.
√
Determine z1 in polar form.
(V).
Obtain an expression for the current flowing through the
inductor, iL (t).
R1
∗ 1.23 If z = 3 − j5, find the value of ln(z).
1.24 If z = 3e jπ /6 , find the value of ez .
i
A
υs(t)
1.25 If z = 3 − j4, find the value of ez .
iL
iR2
+
R2
L
−
Section 1-6: Phasors
R1 = 20 Ω, R2 = 30 Ω, L = 0.4 mH
∗ 1.26 A voltage source given by
υs (t) = 25 cos(2π × 103t − 30◦)
(V)
is connected to a series RC load as shown in Fig. 1-20.
If R = 1 MΩ and C = 200 pF, obtain an expression for the
voltage across the capacitor, υc (t).
Figure P1.30 Circuit for Problem 1.30.
1.31 Repeat Problem 1.30, but this time determine current
iR2 (t).
Chapter
2
Transmission Lines
Chapter Contents
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
TB3
2-9
2-10
2-11
2-12
TB4
Objectives
Upon learning the material presented in this chapter, you
should be able to:
General Considerations, 47
Lumped-Element Model, 50
Transmission-Line Equations, 53
Wave Propagation on a Transmission Line, 54
The Lossless Microstrip Line, 59
The Lossless Transmission Line: General
Considerations, 62
Wave Impedance of the Lossless Line, 71
Special Cases of the Lossless Line, 74
Microwave Ovens, 80
Power Flow on a Lossless Transmission Line, 82
The Smith Chart, 84
Impedance Matching, 94
Transients on Transmission Lines, 108
EM Cancer Zappers, 117
Chapter 2 Summary, 119
Problems, 121
1. Calculate the line parameters, characteristic impedance,
and propagation constant of coaxial, two-wire, parallelplate, and microstrip transmission lines.
2. Determine the reflection coefficient at the load-end of
the transmission line, the standing-wave pattern, and the
locations of voltage and current maxima and minima.
3. Calculate the amount of power transferred from the generator to the load through the transmission line.
4. Use the Smith chart to perform transmission-line calculations.
5. Analyze the response of a transmission line to a voltage
pulse.
46
2-1 GENERAL CONSIDERATIONS
Zg
47
A
B
+
~
Vg
−
Sending-end
port
Receiving-end
port
Transmission line
A'
ZL
B'
Generator circuit
Load circuit
Figure 2-1 A transmission line is a two-port network connecting a generator circuit at the sending end to a load at the receiving end.
2-1 General Considerations
I
n most electrical engineering curricula, the study of electromagnetics is preceded by one or more courses on electrical
circuits. In this book, we use this background to build a
bridge between circuit theory and electromagnetic theory. The
bridge is provided by transmission lines, which is the topic
of this chapter. By modeling transmission lines in the form of
equivalent circuits, we can use Kirchhoff’s voltage and current laws to develop wave equations whose solutions provide
an understanding of wave propagation, standing waves, and
power transfer. Familiarity with these concepts facilitates the
presentation of material in later chapters.
Although the notion of transmission lines may encompass
all structures and media that serve to transfer energy or
information between two points, including nerve fibers in the
human body and fluids and solids that support the propagation of mechanical pressure waves, this chapter focuses on
transmission lines that guide electromagnetic signals. Such
transmission lines include telephone wires, coaxial cables
carrying audio and video information to TV sets or digital
data to computer monitors, microstrips printed on microwave
circuit boards, and optical fibers carrying light waves for the
transmission of data at very high rates.
Fundamentally, a transmission line is a two-port network,
with each port consisting of two terminals, as illustrated in
Fig. 2-1. One of the ports, the line’s sending end, is connected
to a source (also called the generator). The other port, the line’s
receiving end, is connected to a load. The source connected
to the transmission line’s sending end may be any circuit
generating an output voltage, such as a radar transmitter, an
amplifier, or a computer terminal operating in transmission
mode. From circuit theory, a dc source can be represented by a
Thévenin-equivalent generator circuit consisting of a generator voltage Vg in series with a generator resistance Rg , as shown
in Fig. 2-1. In the case of alternating-current (ac) signals, the
eg and an
generator circuit is represented by a voltage phasor V
impedance Zg .
The load circuit, or simply the load, may be a transmitting
antenna in the case of radar, a computer terminal operating in
the receiving mode, the input terminals of an amplifier, or any
output circuit whose input terminals can be represented by an
equivalent load impedance ZL .
2-1.1 The Role of Wavelength
In low-frequency circuits, circuit elements usually are interconnected using simple wires. In the circuit shown in Fig. 2-2,
for example, the generator is connected to a simple RC load
via a pair of wires. In view of our definition in the preceding
paragraphs of what constitutes a transmission line, we pose
the question: Is the pair of wires between terminals AA′ and
terminals BB′ a transmission line? If so, under what set of
circumstances should we explicitly treat the pair of wires
as a transmission line as opposed to ignoring their presence
altogether and treating the circuit as only an RC load connected
to a generator Veg ? The answer to the first question is: Yes,
A
i
+
B
+
+
Vg
R
VAA'
Transmission line
VBB'
−
C
−
A'
−
B'
l
Figure 2-2 Generator connected to an RC circuit through a
transmission line of length l.
48
CHAPTER 2
the pair of wires does constitute a transmission line. The
answer to the second question is: The factors that determine
whether or not we should treat the wires as a transmission line
is governed by the length of the line l and the frequency f
of the signal provided by the generator. (As we will see
later, the determining factor is the ratio of the length l to the
wavelength λ of the wave propagating on the transmission
line between the source at terminals AA′ and the load at
terminals BB′ .) If the generator voltage is cosinusoidal in time,
then the voltage across the input terminals AA′ is
VAA′ = Vg (t) = V0 cos ω t
(V),
(2.1)
where ω = 2π f is the angular frequency. If we assume that the
current flowing through the wires travels at the speed of light,
c = 3 × 108 m/s, then the voltage across the output terminals
BB′ will have to be delayed in time relative to that across AA′
by the travel delay time l/c. Thus, assuming no ohmic losses
in the transmission line and ignoring other transmission-line
effects discussed later in this chapter,
VBB′ (t) = VAA′ (t − l/c) = V0 cos[ω (t − l/c)] = V0 cos(ω t − φ0 ),
(2.2)
with
ωl
(rad).
(2.3)
φ0 =
c
Thus, the time delay associated with the length of the
line l manifests itself as a constant phase shift φ0 in
the argument of the cosine. Let us compare VBB′ with
VAA′ at t = 0 for an ultralow-frequency electronic circuit
operating at a frequency f = 1 kHz. For a typical wire
length l = 5 cm, Eqs. (2.1) and (2.2) give VAA′ = V0 and
VBB′ = V0 cos(2π f l/c) = 0.999999999998V0. Hence, for all
practical purposes, the presence of the transmission line may
be ignored, and terminal AA′ may be treated as identical with
BB′ so far as its voltage is concerned. On the other hand,
had the line been a 20-km long telephone cable carrying a
1 kHz voice signal, then the same calculation would have led
to VBB′ = 0.91V0, which is a deviation of 9%. Furthermore, had
the frequency of the signal on the 5-cm long wire been 1.5 GHz
instead of 1 kHz, the voltage at the end of the line would have
been VBB′ = 0 ! Thus, the pair of wires would have a voltage V0
at AA′ and zero at BB′ simultaneously. The determining factor
is the magnitude of φ0 = ω l/c. From Eq. (1.27), the velocity of
propagation up of a traveling wave is related to the oscillation
frequency f and the wavelength λ by
up = f λ
(m/s).
In the present case, up = c. Hence, the phase delay is
φ0 =
ω l 2π f l
l
=
= 2π
c
c
λ
radians.
(2.4)
TRANSMISSION LINES
◮ When l/λ is very small, transmission-line effects may
be ignored, but when l/λ & 0.01, it may be necessary to
account not only for the phase shift due to the time delay
but also for the presence of reflected signals that may have
been bounced back by the load toward the generator. ◭
Power loss on the line and dispersive effects may need to be
considered as well.
◮ A dispersive transmission line is one on which the wave
velocity is not constant as a function of the frequency f . ◭
This means that the shape of a rectangular pulse, which
through Fourier analysis can be decomposed into many sinusoidal waves of different frequencies, gets distorted as it travels
down the line because its different frequency components do
not propagate at the same velocity (Fig. 2-3). Preservation of
pulse shape is very important in high-speed data transmission
not only between terminals but also across transmission line
segments fabricated within high-speed integrated circuits. At
10 GHz, for example, the wavelength is λ = 3 cm in air but
only on the order of 1 cm in a semiconductor material. Hence,
even lengths between devices on the order of millimeters
become significant, and their presence has to be accounted for
in the design of the circuit.
Dispersionless line
Short dispersive line
Long dispersive line
Figure 2-3 A dispersionless line does not distort signals
passing through it regardless of its length, whereas a dispersive
line distorts the shape of the input pulses because the different
frequency components propagate at different velocities. The
degree of distortion is proportional to the length of the dispersive
line.
2-1 GENERAL CONSIDERATIONS
49
Metal
Metal
w
2b
2a
d
h
D
Dielectric spacing
Dielectric spacing
(a) Coaxial line
Dielectric spacing
(b) Two-wire line
(c) Parallel-plate line
Metal strip conductor
w
Metal
Metal
h
Dielectric spacing
(d) Strip line
Metal ground plane
Dielectric spacing
Metal ground plane
Dielectric spacing
(e) Microstrip line
(f) Coplanar waveguide
TEM Transmission Lines
Metal
Concentric
dielectric
layers
(g) Rectangular waveguide
(h) Optical fiber
Higher-Order Transmission Lines
Figure 2-4 A few examples of transverse electromagnetic (TEM) and higher-order transmission lines.
2-1.2 Propagation Modes
A few examples of common types of transmission lines are
shown in Fig. 2-4. Transmission lines may be classified into
two basic types:
• Transverse electromagnetic (TEM) transmission lines:
Waves propagating along these lines are characterized by
electric and magnetic fields that are entirely transverse
to the direction of propagation. Such an orthogonal configuration is called a TEM mode. A good example is the
coaxial line shown in Fig. 2-5. The electric field is in the
radial direction between the inner and outer conductors,
while the magnetic field circles the inner conductor,
and neither has a component along the line axis (the
direction of wave propagation). Other TEM transmission
lines include the two-wire line and the parallel-plate line
(both shown in Fig. 2-4). Although the fields present on
a microstrip line do not adhere to the exact definition
of a TEM mode, the nontransverse field components
are sufficiently small (in comparison with the transverse
components) that they may be ignored, thereby allowing
the inclusion of microstrip lines in the TEM class. A
common feature among TEM lines is that they consist of
two parallel conducting surfaces.
• Higher-order transmission lines: Waves propagating
along these lines have at least one significant field component in the direction of propagation. Hollow conducting
waveguides, dielectric rods, and optical fibers belong to
this class of lines (Chapter 8).
50
CHAPTER 2
TRANSMISSION LINES
Magnetic field lines
Electric field lines
Rg
+
Vg
RL
Coaxial line
−
Generator
Load
Cross section
Figure 2-5 In a coaxial line, the electric field is in the radial direction between the inner and outer conductors, and the magnetic field
forms circles around the inner conductor. The coaxial line is a transverse electromagnetic (TEM) transmission line because both the electric
and magnetic fields are orthogonal to the direction of propagation between the generator and the load.
Only TEM-mode transmission lines are treated in this chapter. This is because they are more commonly used in practice,
and fortunately, less mathematical rigor is required for treating
them than is required for lines that support higher-order modes.
We start our treatment by representing the transmission line
in terms of a lumped-element circuit model, and then we
apply Kirchhoff’s voltage and current laws to derive a pair of
equations governing their behavior, known as the telegrapher’s
equations. By combining these equations, we obtain wave
equations for the voltage and current at any location along
the line. Solution of the wave equations for the sinusoidal
steady-state case leads to a set of formulas that can be used
to solve a wide range of practical problems. In the latter part
of this chapter, we introduce a graphical tool known as the
Smith chart, which facilitates the solution of transmissionline problems without having to perform laborious calculations
involving complex numbers.
2-2 Lumped-Element Model
When we draw a schematic of an electronic circuit, we use
specific symbols to represent resistors, capacitors, inductors,
diodes, and the like. In each case, the symbol represents the
functionality of the device rather than its shape, size, or other
attributes. We shall do the same for transmission lines.
◮ A transmission line will be represented by a parallelwire configuration (Fig. 2-6(a)), regardless of its specific
shape or constitutive parameters. ◭
Thus, Fig. 2-6(a) may represent a coaxial line, a two-wire line,
or any other TEM line.
Drawing again on our familiarity with electronic circuits,
when we analyze a circuit containing a transistor, we mimic
the functionality of the transistor by an equivalent circuit composed of sources, resistors, and capacitors. We apply the same
approach to the transmission line by orienting the line along
the z direction, subdividing it into differential sections each
of length ∆z (Fig. 2-6(b)), and then representing each section
by an equivalent circuit, as illustrated in Fig. 2-6(c). This
representation, often called the lumped-element circuit model,
consists of four basic elements with values that henceforth will
be called the transmission line parameters. These are:
• R′ : The combined resistance of both conductors per unit
length, in Ω/m,
• L′ : The combined inductance of both conductors per unit
length, in H/m,
• G′ : The conductance of the insulation medium between the
two conductors per unit length, in S/m, and
• C ′ : The capacitance of the two conductors per unit length, in
F/m.
Whereas the four line parameters are characterized by different
formulas for different types of transmission lines, the equivalent model represented by Fig. 2-6(c) is equally applicable to
all TEM transmission lines. The prime superscript is used as
a reminder that the line parameters are differential quantities
whose units are per unit length.
Expressions for the line parameters R′ , L′ , G′ , and C ′ are
given in Table 2-1 for the three types of TEM transmission
lines diagrammed in parts (a) through (c) of Fig. 2-4. For
2-2 LUMPED-ELEMENT MODEL
51
(a) Parallel-wire representation
∆z
∆z
∆z
∆z
(b) Differential sections each ∆z long
R' ∆z
L' ∆z
G' ∆z
R' ∆z
C' ∆z
L' ∆z
G' ∆z
∆z
R' ∆z
C' ∆z
∆z
L' ∆z
G' ∆z
R' ∆z
C' ∆z
∆z
L' ∆z
G' ∆z
C' ∆z
∆z
(c) Each section is represented by an equivalent circuit
Figure 2-6 Regardless of its cross-sectional shape, a TEM transmission line is represented by the parallel-wire configuration shown in
part (a). To obtain equations relating voltages and currents, the line is subdivided into small differential sections in part (b), each of which
is then represented by an equivalent circuit in part (c).
each of these lines, the expressions are functions of two sets
of parameters: (1) geometric parameters defining the crosssectional dimensions of the given line and (2) the electromagnetic constitutive parameters of the conducting and insulating
materials. The pertinent geometric parameters are:
• Coaxial line (Fig. 2-4(a)):
a = outer radius of inner conductor, m
b = inner radius of outer conductor, m
• Two-wire line (Fig. 2-4(b)):
d = diameter of each wire, m
D = spacing between wires’ centers, m
• Parallel-plate line (Fig. 2-4(c)):
w = width of each plate, m
h = thickness of insulation between plates, m
◮ The pertinent constitutive parameters apply to all three
lines and consist of two groups:
(1) µc and σc are the magnetic permeability and electrical conductivity of the conductors, and
(2) ε , µ , and σ are the electrical permittivity, magnetic
permeability, and electrical conductivity of the insulation material separating them. ◭
Appendix B contains tabulated values for these constitutive
parameters for various materials. For the purposes of this chapter, we need not concern ourselves with the derivations leading
to the expressions in Table 2-1. The techniques necessary for
computing R′ , L′ , G′ , and C ′ for the general case of an arbitrary
two-conductor configuration are presented in later chapters.
The lumped-element model shown in Fig. 2-6(c) reflects the
physical phenomena associated with the currents and voltages
on any TEM transmission line. It consists of two in-series
elements, R′ and L′ , and two shunt elements, G′ and C ′ . To
52
CHAPTER 2
TRANSMISSION LINES
Table 2-1 Transmission-line parameters R′ , L′ , G′ , and C ′ for three types of lines.
Parameter
R′
Rs
2π
Coaxial
1 1
+
a b
L′
µ
ln(b/a)
2π
G′
2πσ
ln(b/a)
C′
2πε
ln(b/a)
Two-Wire
Parallel-Plate
Unit
2Rs
πd
2Rs
w
Ω/m
q
µ
2
ln (D/d) + (D/d) − 1
π
µh
w
H/m
σw
h
S/m
εw
h
F/m
πσ
h
i
p
ln (D/d) + (D/d)2 − 1
πε
h
i
p
ln (D/d) + (D/d)2 − 1
Notes: (1) Refer to Fig. 2-4 for definitions of
pdimensions. (2) µ , ε , and σ pertain to the insulating
material between the conductors.
(3) Rs = π f µci/σc . (4) µc and σc pertain to the conductors.
h
p
(5) If (D/d)2 ≫ 1, then ln (D/d) + (D/d)2 − 1 ≈ ln(2D/d).
explain the lumped-element model, consider a small section of
a coaxial line, as shown in Fig. 2-7. The line consists of inner
and outer conductors of radii a and b separated by a material
with permittivity ε , permeability µ , and conductivity σ . The
two metal conductors are made of a material with conductivity σc and permeability µc .
Resistance R′
When a voltage source is connected across the terminals
connected to the two conductors at the sending end of the line,
currents flow through the conductors, primarily along the outer
surface of the inner conductor and the inner surface of the outer
conductor. The line resistance R′ accounts for the combined
resistance per unit length of the inner and outer conductors.
The expression for R′ is derived in Chapter 7 and is given by
Eq. (7.96) as
Rs 1 1
′
(coax line)
(Ω/m),
(2.5)
R =
+
2π a b
where Rs , which represents the surface resistance of the
conductors, is given by Eq. (7.92a) as
r
π f µc
Rs =
(Ω).
(2.6)
σc
The surface resistance depends not only on the material
properties of the conductors (σc and µc ), but also on the
frequency f of the wave traveling on the line.
Conductors
(µc, σc)
b
a
Insulating material
(ε, µ, σ)
Figure 2-7 Cross section of a coaxial line with inner conductor
of radius a and outer conductor of radius b. The conductors
have magnetic permeability µc and conductivity σc , and the
spacing material between the conductors has permittivity ε ,
permeability µ , and conductivity σ .
◮ For a perfect conductor with σc = ∞ or a highconductivity material such that ( f µc /σc ) ≪ 1, Rs
approaches zero, and so does R′ . ◭
2-3 TRANSMISSION-LINE EQUATIONS
53
Inductance L′
Next, let us examine the line inductance L′ , which accounts
for the joint inductance of both conductors. Application of
Ampère’s law in Chapter 5 to the definition of inductance leads
to the following expression [Eq. (5.99)] for the inductance per
unit length of a coaxial line:
b
µ
L′ =
ln
(coax line)
(H/m).
(2.7)
2π
a
Conductance G′
The line conductance G′ accounts for current flow between the
outer and inner conductors, made possible by the conductivity σ of the insulator. It is precisely because the current flow
is from one conductor to the other that G′ appears as a shunt
element in the lumped-element model. For the coaxial line, the
conductance per unit length is given by Eq. (4.86) as
G′ =
2πσ
ln(b/a)
(coax line)
(S/m).
(2.8)
◮ If the material separating the inner and outer conductors
is a perfect dielectric with σ = 0, then G′ = 0. ◭
two-wire air line). For an air line, ε = ε0 = 8.854 × 10−12 F/m,
µ = µ0 = 4π × 10−7 H/m, σ = 0, and G′ = 0.
What is a transmission line?
When should transmission-line effects be considered, and
when may they be ignored?
Concept Question 2-1:
What is the difference between
dispersive and nondispersive transmission lines? What is
the practical significance of dispersion?
Concept Question 2-2:
Concept Question 2-3:
What constitutes a TEM trans-
mission line?
What purpose does the
lumped-element circuit model serve? How are the line
parameters R′ , L′ , G′ , and C ′ related to the physical
and electromagnetic constitutive properties of the
transmission line?
Concept Question 2-4:
Exercise 2-1: Use Table 2-1 to evaluate the line parameters of a two-wire air line with wires of radius 1 mm,
separated by a distance of 2 cm. The wires may be treated
as perfect conductors with σc = ∞.
R′ = 0, L′ = 1.20 (µ H/m),
= 9.29 (pF/m). (See EM .)
Answer:
G′ = 0,
Capacitance C ′
C′
The last line parameter on our list is the line capacitance C ′ .
When equal and opposite charges are placed on any two noncontacting conductors, a voltage difference develops between
them. Capacitance is defined as the ratio of the charge to the
voltage difference. For the coaxial line, the capacitance per
unit length is given by Eq. (4.127) as
Exercise 2-2: Calculate the transmission line parameters
at 1 MHz for a coaxial air line with inner and outer
conductor diameters of 0.6 cm and 1.2 cm, respectively.
The conductors are made of copper (see Appendix B for
µc and σc of copper).
C′ =
2πε
ln(b/a)
(coax line)
(F/m).
(2.9)
All TEM transmission lines share the following useful
relations:
L′C ′ = µε
(all TEM lines),
(2.10)
and
G′
σ
=
C′
ε
(all TEM lines).
(2.11)
If the insulating medium between the conductors is air, the
transmission line is called an air line (e.g., coaxial air line or
Answer: R′ = 2.07 × 10−2 (Ω/m), L′ = 0.14 (µ H/m),
G′ = 0, C ′ = 80.3 (pF/m). (See
EM
.)
2-3 Transmission-Line Equations
A transmission line usually connects a source on one end to
a load on the other. Before considering the complete circuit,
however, we will develop general equations that describe the
voltage across and current carried by the transmission line as a
function of time t and spatial position z. Using the lumpedelement model of Fig. 2-6(c), we begin by considering a
differential length ∆z as shown in Fig. 2-8. The quantities
υ (z,t) and i(z,t) denote the instantaneous voltage and current
at the left end of the differential section (node N), and similarly
υ (z + ∆z,t) and i(z + ∆z,t) denote the same quantities at node
54
CHAPTER 2
Node
N i(z, t)
+
Node
N+1
R' ∆z
υ(z, t)
i(z + ∆z, t)
+
L' ∆z
G' ∆z
C' ∆z
Except for the last section of this chapter, our primary
interest is in sinusoidal steady-state conditions. To that end, we
make use of the phasor representation with a cosine reference,
as outlined in Section 1-7. Thus, we define
υ (z,t) = Re[Ve (z) e jω t ],
˜ e
i(z,t) = Re[I(z)
υ(z + ∆z, t)
−
−
∆z
Figure 2-8 Equivalent circuit of a two-conductor transmission
line of differential length ∆z.
TRANSMISSION LINES
jω t
],
υ (z,t) − R′ ∆z i(z,t) − L′ ∆z
∂ i(z,t)
− υ (z + ∆z,t) = 0. (2.12)
∂t
Upon dividing all terms by ∆z and rearranging them, we obtain
υ (z + ∆z, t) − υ (z,t)
∂ i(z,t)
= R′ i(z,t) + L′
. (2.13)
−
∆z
∂t
In the limit as ∆z → 0, Eq. (2.13) becomes a differential
equation:
−
∂ υ (z,t)
∂ i(z,t)
= R′ i(z,t) + L′
.
∂z
∂t
(2.14)
Similarly, Kirchhoff’s current law accounts for current drawn
from the upper line at node (N + 1) by the parallel conductance
G′ ∆z and capacitance C ′ ∆z:
i(z,t) − G′ ∆z υ (z + ∆z,t)
− C ′ ∆z
∂ υ (z + ∆z,t)
− i(z + ∆z,t) = 0.
∂t
(2.15)
dVe (z)
˜
= (R′ + jω L′ ) I(z),
dz
˜
d I(z)
= (G′ + jω C ′ ) Ve (z).
−
dz
(telegrapher’s equations in phasor form)
∂ i(z,t)
∂ υ (z,t)
= G′ υ (z,t) + C ′
.
∂z
∂t
(2.16)
The first-order differential equations (2.14) and (2.16) are the
time-domain forms of the transmission-line equations, known
as the telegrapher’s equations.
(2.18a)
(2.18b)
2-4 Wave Propagation on a Transmission
Line
The two first-order coupled equations (2.18a) and (2.18b) can
be combined to give two second-order uncoupled wave equae (z) and another for I(z).
˜
tions: one for V
The wave equation for
e
V (z) is derived by first differentiating both sides of Eq. (2.18a)
with respect to z, resulting in
˜
d I(z)
d 2Ve (z)
= (R′ + jω L′ )
.
(2.19)
2
dz
dz
˜
Then, upon substituting Eq. (2.18b) for d I(z)/dz,
Eq. (2.19)
becomes
e (z)
d 2V
− (R′ + jω L′ )(G′ + jω C ′ ) Ve (z) = 0,
(2.20)
dz2
or
−
d 2Ve (z)
e (z) = 0,
− γ2 V
dz2
e (z))
(wave equation for V
Upon dividing all terms by ∆z and taking the limit ∆z → 0,
Eq. (2.15) becomes a second-order differential equation:
−
(2.17b)
e (z) and I(z)
˜ are the phasor counterparts of υ (z,t) and
where V
i(z,t), respectively, each of which may be real or complex.
Upon substituting Eqs. (2.17a) and (2.17b) into Eqs. (2.14) and
(2.16), and utilizing the property given by Eq. (1.62) that ∂ /∂ t
in the time domain is equivalent to multiplication by jω in the
phasor domain, we obtain the following pair of equations:
−
(N + 1), located at the right end of the section. Application of
Kirchhoff’s voltage law accounts for the voltage drop across
the series resistance R′ ∆z and inductance L′ ∆z:
(2.17a)
where
γ=
p
(R′ + jω L′ )(G′ + jω C ′ ) .
(propagation constant)
(2.21)
(2.22)
2-4 WAVE PROPAGATION ON A TRANSMISSION LINE
55
Application of the same steps to Eqs. (2.18a) and (2.18b) in
reverse order leads to
(Np/m),
Figure 2-9 In general, a transmission line can support two
traveling waves: an incident wave (with voltage and current
amplitudes (V0+ , I0+ )) traveling along the +z direction (towards
the load) and a reflected wave (with (V0− , I0− )) traveling along
the −z direction (towards the source).
In their present form, the solutions given by Eqs. (2.26a)
and (2.26b) contain four unknowns: the wave amplitudes
(V0+ , I0+ ) of the +z propagating wave and (V0− , I0− ) of the
−z propagating wave. We can easily relate the current wave
amplitudes, I0+ and I0− , to the voltage wave amplitudes, V0+
and V0− , by using Eq. (2.26a) in Eq. (2.18a) and then solving
˜
for the current I(z).
The process leads to
(2.25a)
˜ =
I(z)
(attenuation constant)
β = Im(γ )
p
= Im
(R′ + jω L′ )(G′ + jω C ′ )
(rad/m).
(2.25b)
γ
[V + e−γ z − V0− eγ z ].
R′ + jω L′ 0
(2.27)
Comparison of each term with the corresponding term in
Eq. (2.26b) leads us to conclude that
V0+
−V0−
,
+ = Z0 =
I0
I0−
(phase constant)
(2.28)
where
2-4.1 Phasor-Domain Solution
In Eqs. (2.25a) and (2.25b), we choose the square-root solutions that give positive values for α and β . For passive transmission lines, α is either zero or positive. Most transmission
lines, and all those considered in this chapter, are of the passive
type. The gain region of a laser is an example of an active
transmission line with a negative α .
The wave equations (2.21) and (2.23) have traveling wave
solutions of the form:
˜
I(z)
ZL
z
with
e (z)
V
(V0+, I0+)e−γz Incident wave
(V0−, I0−)eγz Reflected wave
−
(2.23)
The second-order differential equations (2.21) and (2.23) are
˜
called wave equations for Ve (z) and I(z),
respectively, and γ is
called the complex propagation constant of the transmission
line. As such, γ consists of a real part α , called the attenuation
constant of the line with units of Np/m, and an imaginary
part β , called the phase constant of the line with units of
rad/m. Thus,
γ = α + jβ
(2.24)
= V0+ e−γ z + V0− eγ z
= I0+ e−γ z + I0− eγ z
Zg
Vg
˜
d 2 I(z)
˜ = 0.
− γ 2 I(z)
dz2
e
(wave equation for I(z))
α = Re(γ )
p
= Re
(R′ + jω L′ )(G′ + jω C ′ )
+
(V),
(2.26a)
(A).
(2.26b)
As shown later, the e−γ z term represents a wave propagating
in the +z direction, while the eγ z term represents a wave
propagating in the −z direction (Fig. 2-9). Verification that
these are indeed valid solutions is easily accomplished by
substituting the expressions given by Eqs. (2.26a) and (2.26b),
as well as their second derivatives, into Eqs. (2.21) and (2.23).
R′ + jω L′
Z0 =
=
γ
s
R′ + jω L′
G′ + j ω C ′
(Ω),
(2.29)
is called the characteristic impedance of the line.
◮ It should be noted that Z0 is equal to the ratio of the
voltage amplitude to the current amplitude for each of
the traveling waves individually (with an additional minus
sign in the case of the −z propagating wave), but it is
e (z) to the total
not equal to the ratio of the total voltage V
˜
current I(z) unless one of the two waves is absent. ◭
It seems reasonable that the voltage-to-current ratios of the
two waves V0+ /I0+ and V0− /I0− are both related to the same
quantity, namely Z0 , but it is not immediately obvious as
56
CHAPTER 2
TRANSMISSION LINES
Module 2.1 Two-Wire Line The input data specifies the geometric and electric parameters of a two-wire transmission
line. The output includes the calculated values for the line parameters, characteristic impedance Z0 , and attenuation and phase
constants, as well as plots of Z0 as a function of d and D.
to why one of the ratios is the negative of the other. The
explanation, which is available in more detail in Chapter 7,
is based on a directional rule that specifies the relationships
between the directions of the electric and magnetic fields of a
TEM wave and its direction of propagation. On a transmission
line, the voltage is related to the electric field E and the
current is related to the magnetic field H. To satisfy the
directional rule, reversing the direction of propagation requires
reversal of the direction (or polarity) of I relative to V . Hence,
V0− /I0− = −V0+ /I0+ .
In terms of Z0 , Eq. (2.27) can be cast in the form
˜ =
I(z)
V0+ −γ z V0− γ z
e −
e .
Z0
Z0
(2.30)
2-4 WAVE PROPAGATION ON A TRANSMISSION LINE
57
Module 2.2 Coaxial Line Except for changing the geometric parameters to those of a coaxial transmission line, this
module offers the same output information as Module 2.1.
According to Eq. (2.29), the characteristic impedance Z0 is
determined by the angular frequency ω of the wave traveling
along the line and the four line parameters (R′ , L′ , G′ , and C ′ ).
These, in turn, are determined by the line geometry and
its constitutive parameters. Consequently, the combination of
Eqs. (2.26a) and (2.30) now contains only two unknowns,
namely V0+ and V0− , as opposed to four.
2-4.2 Time-Domain Solution for υ (z,t)
In later sections, we apply boundary conditions at the source
and load ends of the transmission line to obtain expressions for
the remaining wave amplitudes V0+ and V0− . In general, each is
a complex quantity characterized by a magnitude and a phase
angle:
+
V0+ = |V0+ |e jφ ,
(2.31a)
58
CHAPTER 2
−
V0− = |V0− |e jφ .
(2.31b)
After substituting these definitions in Eq. (2.26a) and using
Eq. (2.24) to decompose γ into its real and imaginary parts, we
can convert back to the time domain to obtain an expression
for υ (z,t), the instantaneous voltage on the line:
υ (z,t) = Re(Ve (z)e jω t )
= Re V0+ e−γ z + V0− eγ z e jω t
+
−
= Re[|V0+ |e jφ e jω t e−(α + jβ )z + |V0− |e jφ e jω t e(α + jβ )z].
The final expression for υ (z,t) is then given by
+ |V0− |eα z cos(ω t + β z + φ −).
(2.32)
From our review of waves in Section 1-4, we recognize the
first term in Eq. (2.32) as a wave traveling in the +z direction
(the coefficients of t and z have opposite signs) and the second
term as a wave traveling in the −z direction (the coefficients of
t and z are both positive). Both waves propagate with a phase
velocity up given by Eq. (1.30):
up = f λ =
ω
.
β
To obtain an expression for i(z,t), we insert the expressions
given by Eqs. (2.31) and (2.34) into Eq. (2.30), multiply both
terms by e jω t , and then take the real part:
+
e e jω t ) = Re |V0 | e jφ + e− jφz e jω t e−(α + jβ z)
i(z,t) = Re(I(z)
|Z0 |
|V0− | jφ − − jφz jω t (α + jβ z)
−
e e
,
e e
|Z0 |
which yields
i(z,t) =
υ (z,t) = |V0+ |e−α z cos(ω t − β z + φ +)
(2.33)
Because the wave is guided by the transmission line, λ often
is called the guide wavelength. The factor e−α z accounts for
the attenuation of the +z propagating wave, and the factor eα z
accounts for the attenuation of the −z propagating wave.
◮ The presence of two waves on the line propagating in
opposite directions produces a standing wave. ◭
To gain a physical understanding of what that means, we later
provide a thorough treatment of standing waves in Section
2-6.2.
2-4.3 Time-Domain Solution for i(z,t)
e
In the phasor domain, the expression for I(z)
given by
e (z), except that the current
Eq. (2.30) is similar to that for V
amplitudes are divided by the characteristic impedance of the
transmission line, Z0 , and the current wave traveling in the −z
direction has a minus sign (second term in Eq. (2.30)). Since
Z0 is, in general, a complex quantity (see Eq. (2.29)), we need
to express it as
Z0 = |Z0 |e jφz .
(2.34)
TRANSMISSION LINES
|V0+ | −α z
cos(ω t − β z + φ + − φz )
e
|Z0 |
−
|V0− | α z
e cos(ω t + β z + φ − − φz ).
|Z0 |
(2.35)
Comparison of the expression for i(z,t) with that given by
Eq. (2.32) for υ (z,t) reveals that:
(a) The amplitudes of the current waves are reduced by |Z0 |
when compared with the amplitudes of the voltage waves.
(b) The amplitude of the second term in Eq. (2.35)—
corresponding to the current wave traveling in the negative z direction—has a minus sign.
(c) The reference phase angle in the expression for i(z,t) has
an extra component, namely −φz , when compared with
the reference phase angle of υ (z,t).
Example 2-1: Air Line
An air line is a transmission line in which air separates the
two conductors, which renders G′ = 0 because σ = 0. In
addition, assume that the conductors are made of a material
with high conductivity so that R′ ≈ 0. For an air line with
a characteristic impedance of 50 Ω and a phase constant of
20 rad/m at 700 MHz, find the line inductance L′ and the line
capacitance C ′ .
Solution: The following quantities are given:
Z0 = 50 Ω,
β = 20 rad/m,
f = 700 MHz = 7 × 108 Hz.
With R′ = G′ = 0, Eqs. (2.25b) and (2.29) reduce to
hp
i
√
√
β = Im
( jω L′ )( jω C ′ ) = Im jω L′C ′ = ω L′C ′ ,
2-5 THE LOSSLESS MICROSTRIP LINE
Z0 =
s
jω L′
=
jω C ′
r
L′
.
C′
Conducting
strip (μc , σc)
The ratio of β to Z0 is β /Z0 = ω C ′ , or
C′ =
59
β
20
=
= 9.09 × 10−11 (F/m)
ω Z0 2π × 7 × 108 × 50
Dielectric
insulator
(ε, μ, σ)
w
= 90.9 (pF/m).
h
p
From Z0 = L′ /C ′ , it follows that
L′ = Z02C ′ = (50)2 × 90.9 × 10−12 = 2.27 × 10−7 (H/m)
= 227 (nH/m).
Conducting ground plane (μc , σc)
Exercise 2-3: Verify that Eq. (2.26a) indeed provides a
solution to the wave equation (2.21).
Answer: (See
EM
(a) Longitudinal view
.)
B
Exercise 2-4: A two-wire air line has the following line
parameters: R′ = 0.404 (mΩ/m), L′ = 2.0 (µ H/m), G′ = 0,
and C ′ = 5.56 (pF/m). For operation at 5 kHz, determine
(a) the attenuation constant α , (b) the phase constant β ,
(c) the phase velocity up , and (d) the characteristic impedance Z0 . (See EM .)
E
(b) Cross-sectional view with E and B field lines
10−7
(a) α = 3.37 ×
(Np/m), (b)
β = 1.05 × 10−4 (rad/m), (c) up = 3.0 × 108 (m/s),
(d) Z0 = (600 − j1.9) Ω = 600 −0.18◦ Ω.
Answer:
2-5 The Lossless Microstrip Line
Because its geometry is well suited for fabrication on printed
circuit boards, the microstrip line is the most common interconnect configuration used in RF and microwave circuits. It
consists of a narrow, very thin strip of copper (or another
good conductor) printed on a dielectric substrate overlaying
a ground plane (Fig. 2-10(a)). The presence of charges of
opposite polarity on its two conducting surfaces gives rise to
electric field lines between them (Fig. 2-10(b)). Also, the flow
of current through the conductors (when part of a closed circuit) generates magnetic field loops around them, as illustrated
in Fig. 2-10(b) for the narrow strip. Even though the patterns
of E and B are not everywhere perfectly orthogonal, they
are approximately so in the region between the conductors,
which is where the E and B fields are concentrated the most.
Accordingly, the microstrip line is considered a quasi-TEM
transmission line, which allows us to describe its voltages
and currents in terms of the one-dimensional TEM model of
Section 2-4, namely Eqs. (2.26) through (2.33).
(c) Microwave circuit
Figure 2-10 Microstrip line: (a) longitudinal view, (b)
cross-sectional view, and (c) circuit example. (Courtesy of
Prof. Gabriel Rebeiz, U. California at San Diego.)
The microstrip line has two geometric parameters, the
width of the elevated strip, w, and the thickness (height) of
the dielectric layer, h. We will ignore the thickness of the
conducting strip because it has a negligible influence on the
propagation properties of the microstrip line so long as the strip
thickness is much smaller than the width w, which is almost
60
CHAPTER 2
always the case in practice. Also, we assume the substrate
material to be a perfect dielectric with σ = 0 and the metal strip
and ground plane to be perfect conductors with σc ≈ ∞. These
two assumptions simplify the analysis considerably without
incurring significant error. Finally, we set µ = µ0 , which is
always true for the dielectric materials used in the fabrication
of microstrip lines. These simplifications reduce the number
of geometric and material parameters to three, namely w, h,
and ε .
Electric field lines always start on the conductor carrying
positive charges and end on the conductor carrying negative
charges. For the coaxial, two-wire, and parallel-plate lines
shown in the upper part of Fig. 2-4, the field lines are confined
to the region between the conductors. A characteristic attribute
of such transmission lines is that the phase velocity of a wave
traveling along any one of them is given by
c
up = √ ,
εr
(2.36a)
where c is the velocity of light in free space and εr is the
relative permittivity of the dielectric medium between the
conductors.
2-5.1 Effective Permittivity
and x and y are intermediate variables given by
εr − 0.9 0.05
,
εr + 3
4
s + 3.7 × 10−4s2
y = 1 + 0.02 ln
s4 + 0.43
x = 0.56
2-5.2
10 −xy
εr + 1
εr − 1
1+
,
+
εeff =
2
2
s
∗ D.
(2.38b)
Characteristic Impedance
The characteristic impedance of the microstrip line is given by
)
(
r
6 + (2π − 6)e−t
4
60
(2.39)
Z0 = √ ln
+ 1+ 2 ,
εeff
s
s
with
t=
30.67
s
0.75
.
(2.40)
Figure 2-11 displays plots of Z0 as a function of s for various
types of dielectric materials.
The corresponding line and propagation parameters are
given by
R′ = 0
(because σc = ∞),
(2.41a)
G′ = 0
(because σ = 0),
√
εeff
,
C′ =
Z0 c
(2.41b)
L′ = Z02C ′ ,
(2.41d)
(2.41c)
Z0 (Ω)
150
Microstrip
s = w/h
w = strip width
h = substrate thickness
100
εr = 6
(2.37a)
50
εr = 2.5
εr = 10
0
where s is the width-to-thickness ratio,
w
s= ,
h
(2.38a)
(2.36b)
Methods for calculating the propagation properties of the
microstrip line are quite complicated and beyond the scope
of this text. However, it is possible to use curve-fit approximations to rigorous solutions to arrive at the following set of
expressions:∗
+ 0.05 ln(1 + 1.7 × 10−4s3 ).
In the microstrip line, even though most of the electric field
lines connecting the strip to the ground plane do pass directly
through the dielectric substrate, a few go through both the air
region above the strip and the dielectric layer (Fig. 2-10(b)).
This nonuniform mixture can be accounted for by defining an
effective relative permittivity εeff such that the phase velocity
is given by an expression that resembles Eq. (2.36a), namely
c
up = √ .
εeff
TRANSMISSION LINES
2
4
s
6
8
10
(2.37b)
H. Schrader, Microstrip Circuit Analysis, Prentice Hall, 1995,
pp. 31–32.
Figure 2-11 Plots of Z0 as a function of s for various types of
dielectric materials.
2-5 THE LOSSLESS MICROSTRIP LINE
61
Module 2.3 Lossless Microstrip Line The output panel lists the values of the transmission line parameters and displays
the variation of Z0 and εeff with h and w.
α =0
(because R′ = G′ = 0),
ω √
β=
εeff .
c
(2.41e)
(2.41f)
2-5.3 Design Process
The preceding expressions allow us to compute the values of
Z0 and the other propagation parameters when given values
for εr , h, and w. This is exactly what is needed in order to
analyze a circuit containing a microstrip transmission line. To
perform the reverse process, namely to design a microstrip line
by selecting values for its w and h such that their ratio yields
the required value of Z0 (to satisfy design specifications),
we need to express s in terms of Z0 . The expression for Z0
given by Eq. (2.39) is rather complicated, so inverting it to
obtain an expression for s in terms of Z0 is rather difficult.
An alternative option is to generate a family of curves similar
to those displayed in Fig. 2-11 and to use them to estimate
62
CHAPTER 2
s for a specified value of Z0 . A logical extension of the
graphical approach is to generate curve-fit expressions that
provide high-accuracy estimates of s. The error associated with
the following formulas is less than 2%:
(a) For Z0 ≤ (44 − 2εr) Ω,
(
w
2
s= =
(q − 1) − ln(2q − 1)
h
π
+
0.52
εr − 1
ln(q − 1) + 0.29 −
2εr
εr
with
q=
)
(2.42)
εeff = 6.11,
Z0 = 49.93 Ω.
The calculated value of Z0 is, for all practical purposes, equal
to the value specified in the problem statement.
60π 2
√ ,
Z0 εr
Exercise 2-5: A microstrip transmission line uses a strip
of width w and height h = 1 mm over a substrate of
relative permittivity εr = 4. What should w be so that the
characteristic impedance of the line is Z0 = 50 Ω?
Answer: w = 2.05 mm. (See
(2.43a)
εr + 1 Z0
εr − 1
0.12
+
.
0.23 +
2 60
εr + 1
εr
(2.43b)
EM
.)
2-6 The Lossless Transmission Line:
General Considerations
with
The above expressions presume that εr , the relative permittivity of the dielectric substrate, has already been specified. For
typical substrate materials including Duroid, Teflon, silicon,
and sapphire, εr ranges between 2 and 15.
Microstrip Line
A 50 Ω microstrip line uses a 0.5-mm thick sapphire substrate
with εr = 9. What is the width of its copper strip?
Solution: Since Z0 = 50 > 44 − 18 = 32, we should use
Eq. (2.43):
r
0.12
εr + 1 Z0
εr − 1
0.23 +
×
+
p=
2
60
εr + 1
εr
r
9 + 1 50
0.12
9−1
=
0.23 +
= 2.06,
×
+
2
60
9+1
9
w
8e p
8e2.06
s = = 2p
= 4.12
= 1.056.
h
e −2 e −2
y = 0.99,
t = 12.51,
w
8e p
s = = 2p
,
h
e −2
Example 2-2:
w = sh = 1.056 × 0.5 mm = 0.53 mm.
x = 0.55,
(b) for Z0 ≥ (44 − 2εr) Ω,
p=
Hence,
To check our calculations, we use s = 1.056 to calculate Z0 to
verify that the value we obtained is indeed equal or close to
50 Ω. With εr = 9, Eqs. (2.37a) to (2.40) yield
and
r
TRANSMISSION LINES
According to the preceding section, a transmission line is fully
characterized by two fundamental parameters: its propagation
constant γ and its characteristic impedance Z0 , both of which
are specified by the angular frequency ω and the line parameters R′ , L′ , G′ , and C ′ .
◮ In many practical situations, the transmission line can
be designed to exhibit low ohmic losses by selecting
conductors with very high conductivities and dielectric
materials (separating the conductors) with negligible conductivities. As a result, R′ and G′ assume very small values
such that R′ ≪ ω L′ and G′ ≪ ω C ′ . ◭
These conditions allow us to set R′ = G′ ≈ 0 in Eq. (2.22),
which yields
√
γ = α + jβ = jω L′C ′ ,
(2.44)
which in turn implies that
α =0
(lossless line),
√
β = ω L′C ′
(lossless line).
(2.45a)
(2.45b)
2-6 THE LOSSLESS TRANSMISSION LINE: GENERAL CONSIDERATIONS
For the characteristic impedance, application of the lossless
line conditions to Eq. (2.29) leads to
Z0 =
r
L′
C′
(lossless line),
(2.46)
which now is a real number. Using the lossless line expression
for β [Eq. (2.45b)], we obtain the following expressions for
the guide wavelength λ and the phase velocity up :
2π
2π
= √
,
β
ω L′C ′
ω
1
.
up = = √
β
L′C ′
λ=
(2.48)
(rad/m),
1
up = √
µε
(2.49)
(m/s),
(2.50)
where µ and ε are, respectively, the magnetic permeability
and electrical permittivity of the insulating material separating
the conductors. Materials used for this purpose are usually
characterized by a permeability µ0 = 4π × 10−7 H/m (the
permeability of free space). Also, the permittivity ε is often
specified in terms of the relative permittivity εr defined as
εr = ε /ε0 ,
−12
(2.51)
−9
where ε0 = 8.854 × 10
F/m ≈ (1/36π ) × 10 F/m is
the permittivity of free space (vacuum). Hence, Eq. (2.50)
becomes
1
1
1
c
=√
·√ =√ ,
(2.52)
µ0 εr ε0
µ0 ε0
εr
εr
√
where c = 1/ µ0 ε0 = 3 × 108 m/s is the velocity of light in
free space. If the insulating material between the conductors
is air, then εr = 1 and up = c. In view of Eq. (2.51) and
the relationship between λ and up given by Eq. (2.33), the
wavelength is given by
up = √
λ=
up
c 1
λ0
= √ =√ ,
f
f εr
εr
choice of the type of insulating material used in a transmission
line is dictated not only by its mechanical properties but by its
electrical properties as well.
According to Eq. (2.52), if εr of the insulating material is
independent of f (which usually is the case for commonly used
TEM lines), the same independence applies to up .
◮ If sinusoidal waves of different frequencies travel on a
transmission line with the same phase velocity, the line is
called nondispersive. ◭
(2.47)
Upon using Eq. (2.10), Eqs. (2.45b) and (2.48) may be rewritten as
√
β = ω µε
63
(2.53)
where λ0 = c/ f is the wavelength in air corresponding to a
frequency f . Note that, because both up and λ depend on εr , the
This is an important feature to consider when digital data are
transmitted in the form of pulses. A rectangular pulse or a
series of pulses is composed of many Fourier components with
different frequencies. If the phase velocity is the same for all
frequency components (or at least for the dominant ones), then
the pulse’s shape does not change as it travels down the line.
In contrast, the shape of a pulse propagating in a dispersive
medium becomes progressively distorted, and the pulse length
increases (stretches out) as a function of the distance traveled
in the medium (Fig. 2-3), thereby imposing a limitation on
the maximum data rate (which is related to the length of the
individual pulses and the spacing between adjacent pulses)
that can be transmitted through the medium without loss of
information.
Table 2-2 provides a list of the expressions for γ , Z0 , and
up for the general case of a lossy line and for several types of
lossless lines. The expressions for the lossless lines are based
on the equations for L′ and C ′ given in Table 2-1.
2-6: For a lossless transmission line,
λ = 20.7 cm at 1 GHz. Find εr of the insulating
material.
Exercise
Answer: εr = 2.1. (See
EM
.)
Exercise 2-7: A lossless transmission line uses a dielectric
insulating material with εr = 4. If its line capacitance is
C ′ = 10 (pF/m), find (a) the phase velocity up , (b) the line
inductance L′ , and (c) the characteristic impedance Z0 .
Answer: (a) up = 1.5 × 108 (m/s), (b) L′ = 4.45 (µ H/m),
(c) Z0 = 667.1 Ω. (See
EM
.)
2-6.1 Voltage Reflection Coefficient
With γ = jβ for the lossless line, Eqs. (2.26a) and (2.30) for
the total voltage and current become
Ve (z) = V0+ e− jβ z + V0− e jβ z ,
(2.54a)
64
CHAPTER 2
TRANSMISSION LINES
Table 2-2 Characteristic parameters of transmission lines.
Propagation
Constant
γ = α + jβ
General case
Lossless
(R′ = G′ = 0)
Lossless coaxial
Lossless
two-wire
Lossless
parallel-plate
γ=
p
Phase
Velocity
up
(R′ + jω L′ )(G′ + jωC ′ )
up = ω /β
Characteristic
Impedance
Z0
s
(R′ + jω L′ )
Z0 =
(G′ + jωC ′ )
p
√
α = 0, β = ω εr /c
√
up = c/ εr
Z0 =
√
α = 0, β = ω εr /c
√
up = c/ εr
√
Z0 = (60/ εr ) ln(b/a)
√
α = 0, β = ω εr /c
√
up = c/ εr
√
εr ) p
Z0 = (120/
· ln (D/d) + (D/d)2 − 1
√
Z0 ≈ (120/ εr ) ln(2D/d),
if D ≫ d
√
α = 0, β = ω εr /c
√
up = c/ εr
√
Z0 = (120π / εr ) (h/w)
L′ /C ′
p
√
Notes: (1) µ = µ0 , ε = εr ε0 , c = 1/ µ0 ε0 , and µ0 /ε0 ≈ (120π ) Ω, where εr is the relative permittivity
of insulating material. (2) For coaxial line, a and b are radii of inner and outer conductors. (3) For two-wire
line, d = wire diameter and D = separation between wire centers. (4) For parallel-plate line, w = width of
plate and h = separation between the plates.
˜ =
I(z)
V0+ − jβ z V0− jβ z
−
e
e .
Z0
Z0
(2.54b)
These expressions contain two unknowns, V0+ and V0− . According to Section 1-7.2, an exponential factor of the form
e− jβ z is associated with a wave traveling in the positive z
direction, from the source (sending end) to the load (receiving
end). Accordingly, we refer to it as the incident wave with
V0+ as its voltage amplitude. Similarly, the term containing
V0− e jβ z represents a reflected wave with voltage amplitude V0−
traveling along the negative z direction from the load to the
source.
To determine V0+ and V0− , we need to consider the lossless
transmission line in the context of the complete circuit, including a generator circuit at its input terminals and a load at its
output terminals, as shown in Fig. 2-12. The line of length l is
terminated in an arbitrary load impedance ZL .
◮ For mathematical convenience, the reference of the
spatial coordinate z is chosen such that z = 0 corresponds
to the location of the load, not the generator. ◭
At the sending end, at z = −l, the line is connected to a
sinusoidal voltage source with phasor voltage Veg and internal
impedance Zg . Since z points from the generator to the load,
positive values of z correspond to locations beyond the load;
therefore, they are irrelevant to our circuit. In future sections,
we will find it more convenient to work with a spatial dimension that also starts at the load but whose direction is opposite
of z. We shall call it the distance from the load d and define it
as d = −z, as shown in Fig. 2-12.
eL , and the phasor
The phasor voltage across the load, V
˜
current through it, IL , are related by the load impedance ZL
as
VeL
ZL =
.
(2.55)
I˜L
The voltage VeL is the total voltage on the line Ve (z) given by
˜ given by Eq. (2.54b).
Eq. (2.54a), and I˜L is the total current I(z)
Both are evaluated at z = 0:
e (z=0) = V + + V − ,
VeL = V
0
0
(2.56a)
2-6 THE LOSSLESS TRANSMISSION LINE: GENERAL CONSIDERATIONS
~
Ii
Zg
~
Vg
Transmission line
+
+
~
+
~
Vi
~
VL
Z0
IL
ZL
65
is the normalized load impedance. In many transmission line
problems, we can streamline the necessary computation by
normalizing all impedances in the circuit to the characteristic
impedance Z0 . Normalized impedances are denoted by lowercase letters.
In view of Eq. (2.28), the ratio of the current amplitudes is
−
−
V0−
I0−
=
−
= −Γ.
I0+
V0+
−
Generator
Load
z
z = −l
z=0
d=l
d=0
d
Figure 2-12 Transmission line of length l connected on one
end to a generator circuit and on the other end to a load ZL . The
load is located at z = 0, and the generator terminals are at z = −l.
Coordinate d is defined as d = −z.
V+ V−
e
I˜L = I(z=0)
= 0 − 0 .
Z0
Z0
(2.56b)
◮ We note that whereas the ratio of the voltage amplitudes
is equal to Γ, the ratio of the current amplitudes is equal
to −Γ. ◭
The reflection coefficient Γ is governed by a single parameter,
the normalized load impedance zL . As indicated by Eq. (2.46),
Z0 of a lossless line is a real number. However, ZL is in general
a complex quantity, as in the case of a series RL circuit, for
example, for which ZL = R + jω L. Hence, in general Γ also is
complex and given by
Γ = |Γ|e jθr ,
Using these expressions in Eq. (2.55), we obtain
ZL =
V0+ + V0−
V0+ − V0−
Z0 .
(2.57)
Solving for V0− gives
V0− =
ZL − Z0
V0+ .
ZL + Z0
(2.58)
◮ The ratio of the amplitudes of the reflected and incident
voltage waves at the load is known as the voltage reflection coefficient Γ. ◭
Γ=
V0−
ZL − Z0 ZL /Z0 − 1
=
=
ZL + Z0 ZL /Z0 + 1
V0+
=
zL − 1
zL + 1
(dimensionless),
(2.59)
◮ A load is said to be matched to a transmission line if
ZL = Z0 because then there will be no reflection by the
load (Γ = 0 and V0− = 0). ◭
On the other hand, when the load is an open circuit (ZL = ∞),
Γ = 1 and V0− = V0+ , and when it is a short circuit (ZL = 0),
Γ = −1 and V0− = −V0+ (Table 2-3).
Reflection Coefficient
of a Series RC Load
A 100 Ω transmission line is connected to a load consisting
of a 50 Ω resistor in series with a 10 pF capacitor. Find the
reflection coefficient at the load for a 100 MHz signal.
Solution: The following quantities are given (Fig. 2-13):
RL = 50 Ω,
where
zL =
ZL
Z0
(2.62)
where |Γ| is the magnitude of Γ and θr is its phase angle. Note
that |Γ| ≤ 1.
Example 2-3:
From Eq. (2.58), it follows that
(2.61)
(2.60)
Z0 = 100 Ω,
CL = 10 pF = 10−11 F,
f = 100 MHz = 108 Hz.
66
CHAPTER 2
TRANSMISSION LINES
Table 2-3 Magnitude and phase of the reflection coefficient for various types of loads. The normalized load impedance
zL = ZL /Z0 = (R + jX)/Z0 = r + jx, where r = R/Z0 and x = X/Z0 are the real and imaginary parts of zL , respectively.
Reflection Coefficient Γ = |Γ|e jθr
Load
Z0
ZL = (r + jx)Z0
Z0
Z0
Z0
|Γ|
1/2
(r − 1)2 + x2
(r + 1)2 + x2
tan−1
x
r−1
θr
− tan−1
x
r+1
0 (no reflection)
irrelevant
(short)
1
±180◦ (phase opposition)
Z0
(open)
1
0 (in-phase)
Z0
jX = jω L
1
±180◦ − 2 tan−1 x
−j
ωC
1
±180◦ − 2 tan−1 x
Z0
Transmission line
jX =
From Eq. (2.59), the voltage reflection coefficient is
A
Z0 = 100 Ω
RL
50 Ω
CL
10 pF
zL − 1
zL + 1
0.5 − j1.59 − 1
=
0.5 − j1.59 + 1
−0.5 − j1.59
=
1.5 − j1.59
Γ=
A'
Figure 2-13 RC load (Example 2-3).
=
◦
−1.67e j72.6
◦
2.19e− j46.7
◦
= −0.76e j119.3 .
The normalized load impedance is
zL =
This result may be converted into the form of Eq. (2.62) by
◦
replacing the minus sign with e− j180 . Thus,
ZL
Z0
RL − j/(ω CL )
=
Z0
1
1
50 − j
=
100
2π × 108 × 10−11
= (0.5 − j1.59) Ω.
◦
Γ = 0.76e j119.3 e− j180
= 0.76e− j60.7
◦
◦
= 0.76 −60.7◦ ,
or
|Γ| = 0.76,
θr = −60.7◦.
2-6 THE LOSSLESS TRANSMISSION LINE: GENERAL CONSIDERATIONS
Example 2-4:
|Γ| for Purely Reactive Load
Show that |Γ| = 1 for a lossless line connected to a purely
reactive load.
Solution: The load impedance of a purely reactive load is
ZL = jXL . From Eq. (2.59), the reflection coefficient is
ZL − Z0
ZL + Z0
jXL − Z0
=
jXL + Z0
Γ=
q
−
Z02 + XL2 e− jθ
−(Z0 − jXL )
= −e− j2θ = 1e j(π −2θ ),
= q
=
(Z0 + jXL )
Z02 + XL2 e jθ
These expressions now contain only one (yet to be determined)
unknown, V0+ . Before we proceed to solve for V0+ , however,
let us examine the physical meaning underlying these expressions. We begin by deriving an expression for |Ve (z)|, which
is the magnitude of Ve (z). Upon using Eq. (2.62) in Eq. (2.63a)
and applying the relation |Ve (z)| = [Ve (z) Ve ∗ (z)]1/2 , where Ve ∗ (z)
is the complex conjugate of Ve (z), we have
nh
i
|Ve (z)| = V0+ (e− jβ z + |Γ|e jθr e jβ z )
io1/2
h
· (V0+ )∗ (e jβ z + |Γ|e− jθr e− jβ z )
h
i1/2
= |V0+ | 1 + |Γ|2 + |Γ|(e j(2β z+θr) + e− j(2β z+θr) )
1/2
,
= |V0+ | 1 + |Γ|2 + 2|Γ| cos(2β z + θr )
e jx + e− jx = 2 cosx
|Γ| = 1
θr = π − 2θ = π − 2 tan−1
XL
Z0
.
Exercise 2-8: A 50 Ω lossless transmission line is ter-
minated in a load with impedance ZL = (30 − j200) Ω.
Calculate the voltage reflection coefficient at the load.
Answer: Γ = 0.93 −27.5◦ . (See
EM
.)
Exercise 2-9: A 150 Ω lossless line is terminated in a
capacitor with impedance ZL = − j30 Ω. Calculate Γ.
Answer: Γ = 1 −157.4◦ . (See
EM
.)
Exercise 2-10: Given that the reflection coefficient at the
load is Γ = 0.6 − j0.3, determine the normalized load
impedance zL .
Answer: zL = 2.2 − j2.4. (See
EM
.)
2-6.2 Standing Waves
Using the relation V0− = ΓV0+ in Eqs. (2.54a) and (2.54b)
yields
Ve (z) = V0+ (e− jβ z + Γe jβ z),
˜ =
I(z)
V0+
Z0
(e− jβ z − Γe jβ z).
(2.63a)
(2.63b)
(2.64)
where we have used the identity
where θ = tan−1 XL /Z0 . Hence, for Γ = |Γ|e jθr ,
and
67
(2.65)
for any real quantity x. To express the magnitude of Ve as a
function of d instead of z, we replace z with −d on the righthand side of Eq. (2.64):
1/2
. (2.66)
|Ve (d)| = |V0+ | 1 + |Γ|2 + 2|Γ| cos(2β d − θr )
By applying the same steps to Eq. (2.63b), a similar expression
˜
can be derived for |I(d)|,
which is the magnitude of the current
˜
I(d):
e
|I(d)|
=
|V0+ |
[1 + |Γ|2 − 2|Γ| cos(2β d − θr )]1/2 .
Z0
(2.67)
˜
The variations of |Ve (d)| and |I(d)|
as a function of d, which
is the position on the line relative to the load (at d = 0), are
illustrated in Fig. 2-14 for a line with |V0+ | = 1 V, |Γ| = 0.3,
θr = 30◦, and Z0 = 50 Ω. The sinusoidal patterns are called
standing waves and are caused by the interference of the
two traveling waves. The maximum value of the standingwave pattern of |Ve (d)| corresponds to the position on the
line at which the incident and reflected waves are in-phase
[2β d − θr = 2nπ in Eq. (2.66)]. Therefore, they add constructively to give a value equal to (1 + |Γ|)|V0+ | = 1.3 V. The
minimum value of |Ve (d)| occurs when the two waves interfere
destructively, which occurs when the incident and reflected
waves are in phase-opposition [2β d − θr = (2n + 1)π ]. In this
case, |Ve (d)| = (1 − |Γ|)|V0+ | = 0.7 V.
◮ Whereas the repetition period is λ for the incident
and reflected waves considered individually, the repetition
period of the standing-wave pattern is λ /2. ◭
68
CHAPTER 2
Voltage
min
~
|V|max
max
λ
dmin λ
3λ
λ
4
2
4
~
(a) |V(d)| versus d
~
|V(d)|
~
|V(d)|
~
|V|min
d
TRANSMISSION LINES
dmax
1.4 V
1.2
1.0
0.8
0.6
0.4
0.2
0
Matched line
|V0+|
d
λ
3λ
4
λ
2
(a) ZL = Z0
λ
4
Short-circuited line
λ/2
0
~
|V(d)|
2|V0+|
~
|I(z)|
Current
~
|I|max
~
|I|min
max
d
λ
3λ
λ
λ
4
2
4
~
(b) |I(d)| versus d
min
30 mA
25
20
15
10
5
0
d
λ
3λ
λ
λ
4
2
4
(b) ZL = 0 (short circuit)
Open-circuited line
0
~
|V(d)|
λ/2
2|V0+|
for a lossless transmission line of characteristic impedance
Z0 = 50 Ω, terminated in a load with a reflection coefficient
◦
Γ = 0.3e j30 . The magnitude of the incident wave |V0+ | = 1 V.
The standing-wave ratio is S = |Ve |max /|Ve |min = 1.3/0.7 = 1.86.
The standing-wave pattern describes the spatial variation of the
magnitude of Ve (d) as a function of d. If one were to observe
the variation of the instantaneous voltage as a function of time
at location d = dmax in Fig. 2-14, that variation would be
as cos ω t and would have an amplitude equal to 1.3 V [i.e.,
υ (t) would oscillate between −1.3 V and +1.3 V]. Similarly,
the instantaneous voltage υ (d,t) at any location d will be
sinusoidal with amplitude equal to |Ve (d)| at that d.
◮ Interactive Module 2.4a provides a highly recommended simulation tool for gaining better understanding
e (d) and I(d)
e and the
of the standing-wave patterns for V
dynamic behavior of υ (d,t) and i(d,t). ◭
a At
λ
3λ
4
(c) ZL =
8
d
˜
Figure 2-14 Standing-wave pattern for (a) |Ve (d)| and (b) |I(d)|
λ
λ
2
4
(open circuit)
0
Figure 2-15 Voltage standing-wave patterns for (a) a matched
load, (b) a short-circuited line, and (c) an open-circuited line.
in phase opposition (when one is at a maximum, the other is
at a minimum, and vice versa). This is a consequence of the
fact that the third term in Eq. (2.66) is preceded by a plus sign,
whereas the third term in Eq. (2.67) is preceded by a minus
sign.
The standing-wave patterns shown in Fig. 2-14 are for a line
◦
with Γ = 0.3 e j30 . The peak-to-peak variation of the pattern
(|Ve |min = (1 − |Γ|)|V0+ | to |Ve |max = (1 + |Γ|)|V0+ |) depends
on |Γ|. For the special case of a matched line with ZL = Z0 , we
have |Γ| = 0 and |Ve (d)| = |V0+ | for all values of d, as shown in
Fig. 2-15(a).
em8e.eecs.umich.edu
Close inspection of the voltage and current standing-wave
patterns shown in Fig. 2-14 reveals that the two patterns are
◮ With no reflected wave present, there is no interference
and no standing waves. ◭
2-6 THE LOSSLESS TRANSMISSION LINE: GENERAL CONSIDERATIONS
The other end of the |Γ| scale, at |Γ| = 1, corresponds to
when the load is a short circuit (Γ = −1) or an open circuit
(Γ = 1). The standing-wave patterns for those two cases are
shown in Figs. 2-15(b) and (c); both exhibit maxima of 2|V0+ |
and minima equal to zero, but the two patterns are spatially
shifted relative to each other by a distance of λ /4. A purely
reactive load (capacitor or inductor) also satisfies the condition
|Γ| = 1, but θr is generally neither zero nor 180◦ (Table 2-3).
Exercise 2.9 examines the standing-wave pattern for a lossless
line terminated in an inductor.
Now let us examine the maximum and minimum values of
the voltage magnitude. From Eq. (2.66), |Ve (d)| is a maximum
when the argument of the cosine function is equal to zero or
a multiple of 2π . Let us denote dmax as the distance from the
load at which |Ve (d)| is a maximum. It then follows that
|Ve (d)| = |Ve |max = |V0+ |[1 + |Γ|],
when
(2.68)
2β dmax − θr = 2nπ ,
(2.69)
θr + 2nπ θr λ nλ
,
=
+
2β
4π
2
(2.70)
with n = 0 or a positive integer. Solving Eq. (2.69) for dmax ,
we have
dmax =
n = 1, 2, . . .
n = 0, 1, 2, . . .
where we have used β = 2π /λ . The phase angle of the
voltage reflection coefficient, θr , is bounded between −π
and π radians. If θr ≥ 0, the first voltage maximum
occurs at dmax = θr λ /4π , corresponding to n = 0, but if
θr < 0, the first physically meaningful maximum occurs at
dmax = (θr λ /4π ) + λ /2, corresponding to n = 1. Negative
values of dmax correspond to locations past the end of the line;
therefore, they have no physical significance.
Similarly, the minima of |Ve (d)| occur at distances dmin
where the argument of the cosine function in Eq. (2.66) is
equal to (2n + 1)π , which gives the result
dmax + λ /4, if dmax < λ /4,
dmax − λ /4, if dmax ≥ λ /4.
S=
1 + |Γ|
|Ve |max
=
1 − |Γ|
|Ve |min
(dimensionless).
(2.73)
This quantity is often referred to by its acronym, VSWR or
the shorter acronym SWR, and it provides a measure of the
mismatch between the load and the transmission line:
◮ For a matched load with Γ = 0, we get S = 1, and for a
line with |Γ| = 1, S = ∞. ◭
The attenuation constant α represents ohmic losses. In view of the model given in
Fig. 2-6(c), what should R′ and G′ be in order to have
no losses? Verify your expectation through the expression
for α given by Eq. (2.25a).
How is the wavelength λ of the
wave traveling on the transmission line related to the freespace wavelength λ0 ?
Concept Question 2-6:
When is a load matched to a
transmission line? Why is it important?
Concept Question 2-7:
What is a standing-wave pattern? Why is its period λ /2 and not λ ?
Concept Question 2-8:
with −π ≤ θr ≤ π . The first minimum corresponds to n = 0.
The spacing between a maximum dmax and the adjacent
minimum dmin is λ /4. Hence, the first minimum occurs at
The ratio of |Ve |max to |Ve |min is called the voltage standingwave ratio S, which from Eqs. (2.68) and (2.71) is given by
(2.71)
when (2β dmin − θr ) = (2n + 1)π ,
dmin =
◮ The locations on the line corresponding to voltage
maxima correspond to current minima, and vice versa. ◭
Concept Question 2-5:
if θr < 0,
if θr ≥ 0,
|Ve |min = |V0+ |[1 − |Γ|],
69
What is the separation between
the location of a voltage maximum and the adjacent
current maximum on the line?
Concept Question 2-9:
(2.72)
70
CHAPTER 2
TRANSMISSION LINES
Module 2.4 Transmission Line Simulator Upon specifying the requisite input data—including the load impedance
at d = 0 and the generator voltage and impedance at d = l, this module provides a wealth of output information about the
voltage and current waveforms along the transmission line. You can view plots of the standing-wave patterns for voltage and
current, the time and spatial variations of the instantaneous voltage υ (d,t) and current i(d,t), and other related quantities.
Exercise 2-11: Use Module 2.4 to generate the volt-
age and current standing-wave patterns for a 50 Ω
line of length 1.5λ terminated in an inductance with
ZL = j140 Ω.
Answer: See Module 2.4 display.
Example 2-5: Standing-Wave Ratio
A 50 Ω transmission line is terminated in a load with
ZL = (100 + j50) Ω. Find the voltage reflection coefficient
and the voltage standing-wave ratio.
Solution: From Eq. (2.59), Γ is given by
Γ=
zL − 1 (2 + j1) − 1 1 + j1
=
=
.
zL + 1 (2 + j1) + 1 3 + j1
Converting the numerator and denominator to polar form
yields
◦
1.414e j45
j26.6◦
Γ=
.
◦ = 0.45e
j18.4
3.162e
Using the definition for S given by Eq. (2.73), we have
S=
1 + |Γ| 1 + 0.45
=
= 2.6.
1 − |Γ| 1 − 0.45
2-7 WAVE IMPEDANCE OF THE LOSSLESS LINE
Example 2-6:
71
Hence,
Measuring ZL
◦
A slotted-line probe is an instrument used to measure the unknown impedance of a load, ZL . A coaxial slotted line contains
a narrow longitudinal slit in the outer conductor. A small probe
inserted in the slit can be used to sample the magnitude of the
electric field and, hence, the magnitude |Ve (d)| of the voltage
on the line (Fig. 2-16). By moving the probe along the length
of the slotted line, it is possible to measure |Ve |max and |Ve |min
and the distances from the load at which they occur. Use of
Eq. (2.73), namely S = |Ve |max /|Ve |min , provides the voltage
standing-wave ratio S. Measurements with a Z = 50 Ω slotted
line terminated in an unknown load impedance determined that
S = 3. The distance between successive voltage minima was
found to be 30 cm, and the first voltage minimum was located
at 12 cm from the load. Determine the load impedance ZL .
Sliding probe
To detector
Probe tip
~+
Vg
−
Slit
Γ = |Γ|e jθr = 0.5e− j36 = 0.405 − j0.294.
Solving Eq. (2.59) for ZL , we have
1 + 0.405 − j0.294
1+Γ
= 50
= (85 − j67) Ω.
ZL = Z0
1−Γ
1 − 0.405 + j0.294
Exercise 2-12: If Γ = 0.5 −60◦ and λ = 24 cm, find the
locations of the voltage maximum and minimum nearest
to the load.
Answer: dmax = 10 cm, dmin = 4 cm. (See
EM
.)
Exercise 2-13: A 140 Ω lossless line is terminated in a
load impedance ZL = (280 + j182) Ω. If λ = 72 cm, find
(a) the reflection coefficient Γ, (b) the voltage standingwave ratio S, (c) the locations of voltage maxima, and (d)
the locations of voltage minima.
(a) Γ = 0.5 29◦ , (b) S = 3.0, (c) dmax =
2.9 cm + nλ /2, (d) dmin = 20.9 cm + nλ /2, where
n = 0, 1, 2, . . . . (See EM .)
Answer:
Zg
40 cm
30 cm
20 cm
10 cm
ZL
Figure 2-16 Slotted coaxial line (Example 2-6).
Solution: The following quantities are given:
Z0 = 50 Ω,
S = 3,
dmin = 12 cm.
Since the distance between successive voltage minima is λ /2,
λ = 2 × 0.3 = 0.6 m,
and
10π
2π
2π
=
(rad/m).
=
λ
0.6
3
From Eq. (2.73), solving for |Γ| in terms of S gives
β=
|Γ| =
S−1 3−1
=
= 0.5.
S+1 3+1
Next, we use the condition given by Eq. (2.71) to find θr :
2β dmin − θr = π ,
for n = 0 (first minimum),
which gives
10π
θr = 2β dmin − π = 2 ×
× 0.12 − π = −0.2π (rad) = −36◦ .
3
2-7 Wave Impedance of the Lossless
Line
The standing-wave patterns indicate that on a mismatched
line the voltage and current magnitudes are oscillatory with
position along the line and in phase opposition with each other.
Hence, the voltage to current ratio, called the wave impedance
Z(d), must vary with position also. Using Eqs. (2.63a) and
(2.63b) with z = −d,
Z(d) =
e (d) V + [e jβ d + Γe− jβ d ]
V
Z0
= 0+ jβ d
˜
I(d)
V0 [e − Γe− jβ d ]
#
"
1 + Γd
1 + Γe− j2β d
= Z0
= Z0
1 − Γd
1 − Γe− j2β d
(Ω),
(2.74)
where we define
Γd = Γe− j2β d = |Γ|e jθr e− j2β d = |Γ|e j(θr −2β d)
(2.75)
as the phase-shifted voltage reflection coefficient, meaning
that Γd has the same magnitude as Γ, but its phase is shifted by
2β d relative to that of Γ.
72
CHAPTER 2
+
Zg
~
Vg
−
A
B
C
~
Ii
Zg
Z0
Z(d )
A′
B′
d=l
d
(a) Actual circuit
TRANSMISSION LINES
ZL
C′
~
Vg
0
+
A
+
Transmission line
~
Vi Zin
~
Z0
VL
+
~
Vg
−
A
B
A′
Z(d )
B′
IL
ZL
−
− A′
−
z = −l
d=l
z=0
d=0
Generator
Zg
~
+
Load
~
Ii
Zg
(b) Equivalent circuit
Figure 2-17 The segment to the right of terminals BB′ can be
replaced with a discrete impedance equal to the wave impedance
Z(d).
~
Vg
+
A
+
~
Vi
Zin
−
− A′
Figure 2-18 At the generator end, the terminated transmission
◮ Z(d) is the ratio of the total voltage (incident- and
reflected-wave voltages) to the total current at any
point d on the line, in contrast with the characteristic impedance of the line Z0 , which relates the voltage and current of each of the two waves individually
(Z0 = V0+ /I0+ = −V0− /I0− ). ◭
In the circuit of Fig. 2-17(a) at terminals BB′ at an arbitrary
location d on the line, Z(d) is the wave impedance of the line
when “looking” to the right, i.e., towards the load. Application
of the equivalence principle allows us to replace the segment
to the right of terminals BB′ with a lumped impedance of
value Z(d), as depicted in Fig. 2-17(b). From the standpoint
of the input circuit to the left of terminals BB′ , the two circuit
configurations are electrically identical.
Of particular interest in many transmission-line problems is
the input impedance at the source end of the line where d = l,
which is given by
1 + Γl
Zin = Z(d = l) = Z0
(2.76)
1 − Γl
with
Γl = Γe− j2β l = |Γ|e j(θr −2β l) .
(2.77)
e jβ l = cos β l + j sin β l,
(2.78a)
By replacing Γ with Eq. (2.59) and using the relations
line can be replaced with the input impedance of the line Zin .
e− jβ l = cos β l − j sin β l,
(2.78b)
Eq. (2.76) can be written in terms of the normalized load
impedance zL as
zL cos β l + j sin β l
Zin = Z0
cos β l + jzL sin β l
zL + j tan β l
= Z0
.
1 + jzL tan β l
(2.79)
From the standpoint of the generator circuit, the transmission
line can be replaced with an impedance Zin , as shown in
Fig. 2-18. The phasor voltage across Zin is given by
eg Zin
V
Vei = I˜i Zin =
.
Zg + Zin
(2.80)
Simultaneously, from the standpoint of the transmission line,
the voltage across it at the input of the line is given by
Eq. (2.63a) with z = −l:
e (−l) = V + [e jβ l + Γe− jβ l ].
Vei = V
0
(2.81)
2-7 WAVE IMPEDANCE OF THE LOSSLESS LINE
73
Equating Eq. (2.80) to Eq. (2.81) and then solving for V0+
leads to
V0+
=
eg Zin
V
Zg + Zin
!
1
.
e jβ l + Γe− jβ l
(2.82)
This completes the solution of the transmission-line wave
equations given by Eqs. (2.21) and (2.23) for the special case
of a lossless transmission line. We started out with the general
solutions given by Eq. (2.26), which included four unknown
amplitudes, V0+ ,V0− , I0+ , and I0− . We then determined that
Z0 = V0+ /I0+ = −V0− /I0− , thereby reducing the unknowns to
the two voltage amplitudes only. Upon applying the boundary
condition at the load, we were able to relate V0− to V0+
through Γ, and finally, by applying the boundary condition at
the source, we obtained an expression for V0+ .
Example 2-7:
Complete Solution for
υ (d,t) and i(d,t)
A 1.05 GHz generator circuit with series impedance Zg = 10 Ω
and voltage source given by
υg (t) = 10 sin(ω t + 30◦)
(V)
is connected to a load ZL = (100 + j50) Ω through a 50-Ω,
67-cm long lossless transmission line. The phase velocity of
the line is 0.7c, where c is the velocity of light in a vacuum.
Find υ (d,t) and i(d,t) on the line.
Solution: From the relationship up = λ f , we find the wavelength
λ=
up
0.7 × 3 × 108
= 0.2 m,
=
f
1.05 × 109
and
βl =
2π
2π
l=
× 0.67 = 6.7π = 0.7π = 126◦ ,
λ
0.2
where we have subtracted multiples of 2π . The voltage reflection coefficient at the load is
◦
ZL − Z0
(100 + j50) − 50
Γ=
=
= 0.45e j26.6 .
ZL + Z0
(100 + j50) + 50
With reference to Fig. 2-18, the input impedance of the line
given by Eq. (2.76) is
1 + Γl
Zin = Z0
1 − Γl
!
1 + Γe− j2β l
= Z0
1 − Γe− j2β l
◦
◦
1 + 0.45e j26.6 e− j252
= (21.9 + j17.4) Ω.
= 50
◦
◦
1 − 0.45e j26.6 e− j252
Rewriting the expression for the generator voltage with the
cosine reference, we have
υg (t) = 10 sin(ω t + 30◦)
= 10 cos(90◦ − ω t − 30◦)
= 10 cos(ω t − 60◦)
◦
= Re[10e− j60 e jω t ] = Re[Veg e jω t ]
eg is given by
Hence, the phasor voltage V
eg = 10 e− j60◦ = 10 −60◦
V
(V).
(V).
Application of Eq. (2.82) gives
!
Veg Zin
1
+
V0 =
Zg + Zin
e jβ l + Γe− jβ l
◦
10e− j60 (21.9 + j17.4)
=
10 + 21.9 + j17.4
◦
◦
◦
· (e j126 + 0.45e j26.6 e− j126 )−1
= 10.2e j159
◦
(V).
Using Eq. (2.63a) with z = −d, the phasor voltage on the line is
Ve (d) = V0+ (e jβ d + Γe− jβ d )
◦
◦
= 10.2e j159 (e jβ d + 0.45e j26.6 e− jβ d ),
and the corresponding instantaneous voltage υ (d,t) is
υ (d,t) = Re[Ve (d) e jω t ]
= 10.2 cos(ω t + β d + 159◦)
+ 4.55 cos(ω t − β d + 185.6◦)
(V).
Similarly, Eq. (2.63b) leads to
˜ = 0.20e j159◦ (e jβ d − 0.45e j26.6◦ e− jβ d ),
I(d)
i(d,t) = 0.20 cos(ω t + β d + 159◦)
+ 0.091 cos(ω t − β d + 185.6◦)
(A).
74
CHAPTER 2
TRANSMISSION LINES
e
Module 2.5 Wave and Input Impedance The wave impedance, Z(d) = Ve (d)/I(d),
exhibits a cyclical pattern as a
function of position along the line. This module displays plots of the real and imaginary parts of Z(d), specifies the locations
of the voltage maximum and minimum nearest to the load, and provides other related information.
2-8 Special Cases of the Lossless Line
We often encounter situations involving lossless transmission
lines with particular terminations or lines whose lengths lead
to particularly useful line properties. We now consider some of
these special cases.
2-8.1
Short-Circuited Line
The transmission line shown in Fig. 2-19(a) is terminated in
a short circuit, ZL = 0. Consequently, the voltage reflection
coefficient defined by Eq. (2.59) is Γ = −1, and the voltage
standing-wave ratio given by Eq. (2.73) is S = ∞. With z = −d
and Γ = −1 in Eqs. (2.63a) and (2.63b), and Γ = −1 in
Eq. (2.74), the voltage, current, and wave impedance on a
short-circuited lossless transmission line are given by
2-8 SPECIAL CASES OF THE LOSSLESS LINE
75
Zsc (d) =
Zinsc
Short
circuit
Z0
d
l
0
(a)
~
Vsc(d)
2jV0+
1
Voltage
d
λ
3λ
4
λ
2
Zinsc =
-1
(b)
~
Isc(d) Z0
2V0+
1
Current
d
λ
3λ
4
λ
2
(2.83c)
esc (d) is zero at the load (d = 0), as it should
The voltage V
be for a short circuit, and its amplitude varies as sin β d. In
contrast, the current I˜sc (d) is a maximum at the load, and it
varies as cos β d. Both quantities are displayed in Fig. 2-19 as
a function of d.
Denoting Zinsc as the input impedance of a short-circuited
line of length l,
0
λ
4
Vesc (d)
= jZ0 tan β d.
Iesc (d)
esc (l)
V
= jZ0 tan β l.
I˜sc (l)
(2.84)
A plot of Zinsc / jZ0 versus l is shown in Fig. 2-19(d). For the
short-circuited line, if its length is less than λ /4, its impedance
is equivalent to that of an inductor, and if it is between λ /4 and
λ /2, it is equivalent to that of a capacitor.
In general, the input impedance Zin of a line terminated in
an arbitrary load has a real part, called the input resistance Rin ,
and an imaginary part, called the input reactance Xin :
0
Zin = Rin + jXin .
-1
In the case of the short-circuited lossless line, the input impedance is purely reactive (Rin = 0). If tan β l ≥ 0, the line appears
inductive to the source, acting like an equivalent inductor Leq
whose impedance equals Zinsc . Thus,
λ
4
(c)
Zinsc
jZ0
Impedance
jω Leq = jZ0 tan β l,
(2.85)
if tan β l ≥ 0,
(2.86)
or
Z0 tan β l
(H).
(2.87)
ω
The minimum line length l that would result in an input impedance Zinsc equivalent to that of an inductor with inductance
Leq is
ω Leq
1
(m).
(2.88)
l = tan−1
β
Z0
Leq =
l
0
λ
3λ
4
λ
2
λ
4
(d)
Figure 2-19 Transmission line terminated in a short circuit: (a)
schematic representation, (b) normalized voltage on the line, (c)
normalized current, and (d) normalized input impedance.
Similarly, if tan β l ≤ 0, the input impedance is capacitive,
in which case the line acts like an equivalent capacitor with
capacitance Ceq such that
1
= jZ0 tan β l,
jω Ceq
or
esc (d) = V + [e jβ d − e− jβ d ] = 2 jV + sin β d,
V
0
0
I˜sc (d) =
V0+ jβ d
2V +
[e + e− jβ d ] = 0 cos β d,
Z0
Z0
(2.83a)
(2.83b)
Ceq = −
if tan β l ≤ 0,
1
Z0 ω tan β l
(F).
(2.89)
(2.90)
Since l is a positive number, the shortest length l for which
tan β l ≤ 0 corresponds to the range π /2 ≤ β l ≤ π . Hence, the
76
CHAPTER 2
minimum line length l that would result in an input impedance
Zinsc equivalent to that of a capacitor of capacitance Ceq is
1
1
−1
l=
π − tan
(m).
(2.91)
β
ω Ceq Z0
◮ These results imply that, through proper choice of
the length of a short-circuited line, we can make them
into equivalent capacitors and inductors of any desired
reactance. ◭
Such a practice is indeed common in the design of microwave
circuits and high-speed integrated circuits, because making an
actual capacitor or inductor often is much more difficult than
fabricating a microstrip transmission line with a shorted end
on a circuit board.
Example 2-8:
Equivalent Reactive Elements
Choose the length of a shorted 50 Ω lossless transmission
line (Fig. 2-20) such that its input impedance at 2.25 GHz is
identical to that of a capacitor with capacitance Ceq = 4 pF.
The wave velocity on the line is 0.75c.
TRANSMISSION LINES
The phase constant is
β=
2π × 2.25 × 109
2π
2π f
=
= 62.8
=
λ
up
2.25 × 108
(rad/m).
From Eq. (2.89), it follows that
tan β l = −
=−
1
Z0 ω Ceq
1
= −0.354.
50 × 2π × 2.25 × 109 × 4 × 10−12
The tangent function is negative when its argument is in
the second or fourth quadrants. The solution for the second
quadrant is
β l1 = 2.8 rad or l1 =
2.8
2.8
= 4.46 cm,
=
β
62.8
and the solution for the fourth quadrant is
β l2 = 5.94 rad or l2 =
5.94
= 9.46 cm.
62.8
We also could have obtained the value of l1 by applying
Eq. (2.91). The length l2 is greater than l1 by exactly λ /2.
In fact, any length l = 4.46 cm + nλ /2, where n is a positive
integer, also is a solution.
l
Zinsc
Z0
Short
circuit
1
Zc =
jωCeq
Zinsc
Figure 2-20
2-8.2
With ZL = ∞, as illustrated in Fig. 2-21(a), we have Γ = 1
and S = ∞, and the voltage, current, and input impedance are
given by
Veoc (d) = V0+ [e jβ d + e− jβ d ] = 2V0+ cos β d,
I˜oc (d) =
(Example 2-8).
up = 0.75c = 0.75 × 3 × 108 = 2.25 × 108 m/s,
Z0 = 50 Ω,
f = 2.25 GHz = 2.25 × 109 Hz,
Ceq = 4 pF = 4 × 10−12 F.
2 jV0+
V0+ jβ d
[e − e− jβ d ] =
sin β d,
Z0
Z0
Zinoc =
Shorted line as an equivalent capacitor
Solution: We are given
Open-Circuited Line
Veoc (l)
= − jZ0 cot β l.
I˜oc (l)
(2.92a)
(2.92b)
(2.93)
Plots of these quantities are displayed in Fig. 2-21 as a function
of d for the voltage and current and as a function of l for the
input impedance.
2-8.3
Application of Short-Circuit/Open-Circuit
Technique
A network analyzer is a radio-frequency (RF) instrument
capable of measuring the impedance of any load connected
2-8 SPECIAL CASES OF THE LOSSLESS LINE
77
gives
Zinoc
d
(a)
Z0
Z0 =
l
d
(b)
λ
3λ
4
λ
2
~
Voc(d)
2V0+
1
λ
4
(2.94)
and the ratio of the same expressions leads to
0
Voltage
p
+
Zinsc Zinoc ,
0
tan β l =
s
−Zinsc
.
Zinoc
(2.95)
Because of the π phase ambiguity associated with the tangent
function, the length l should be less than or equal to λ /2 to
provide an unambiguous result.
−1
~
Current
d
(c)
λ
3λ
4
λ
2
λ
4
Ioc(d) Z0
2jV0+
1
0
Example 2-9: Measuring Z0 and β
Find Z0 and β of a 57-cm long lossless transmission line
whose input impedance was measured as Zinsc = j40.42 Ω when
terminated in a short circuit and as Zinoc = − j121.24 Ω when
terminated in an open circuit. From other measurements, we
know that the line is between 3 and 3.25 wavelengths long.
Solution: From Eqs. (2.94) and (2.95),
−1
p
p
+
Zinsc Zinoc = ( j40.42)(− j121.24) = 70 Ω,
s
r
−Zinsc
1
tan β l =
=
.
oc
Zin
3
Z0 =
Impedance
l
(d)
λ
3λ
4
λ
2
λ
4
Zinoc
jZ0
0
Since l is between 3λ and 3.25λ , β l = (2π l/λ ) is between
6π radians and (13π /2) radians. This places β l in the first
quadrant (0 to π /2)pradians. Hence, the only acceptable
solution for tan β ℓ = 1/3 is β l = π /6 radians. This value,
however, does not include the 2π multiples associated with the
integer λ multiples of l. Hence, the true value of β l is
Figure 2-21 Transmission line terminated in an open circuit:
(a) schematic representation, (b) normalized voltage on the line,
(c) normalized current, and (d) normalized input impedance.
β l = 6π +
π
= 19.4
6
(rad),
in which case
to its input terminal. When used to measure (1) Zinsc , which is
the input impedance of a lossless line terminated in a short
circuit, and (2) Zinoc , which is the input impedance of the line
when terminated in an open circuit, the combination of the
two measurements can be used to determine the characteristic
impedance of the line Z0 and its phase constant β . Indeed,
the product of the expressions given by Eqs. (2.84) and (2.93)
β=
19.4
= 34
0.57
(rad/m).
2-8.4 Lines of Length l = nλ /2
If l = nλ /2, where n is an integer,
tan β l = tan [(2π /λ )(nλ /2)] = tan nπ = 0.
78
CHAPTER 2
Consequently, Eq. (2.79) reduces to
or
Z02 =
for l = nλ /2,
Zin = ZL ,
(2.96)
which means that a half-wavelength line (or any integer multiple of λ /2) does not modify the load impedance.
2-8.5 Quarter-Wavelength Transformer
Another case of interest is when the length of the line is a
quarter-wavelength (or λ /4 + nλ /2, where n = 0 or a positive
integer), corresponding to β l = (2π /λ )(λ /4) = π /2. From
Eq. (2.79), the input impedance becomes
Zin =
Z02
,
ZL
for l = λ /4 + nλ /2.
(2.97)
The utility of such a quarter-wave transformer is illustrated by
Example 2-10.
λ /4 Transformer
Example 2-10:
TRANSMISSION LINES
√
√
Zin ZL = 50 × 100 = 70.7 Ω.
Whereas this eliminates reflections on the feedline, it does not
eliminate them on the λ /4 line. However, since the lines are
lossless, all of the power incident on AA′ will end up getting
transferred into the load ZL .
◮ In this example, ZL is purely resistive. To apply the
λ /4 transformer technique to match a transmission line
to a load with a complex impedance, a more elaborate
procedure is required (Section 2-11). ◭
2-8.6
Matched Transmission Line: ZL = Z0
For a matched lossless transmission line with ZL = Z0 , (1) the
input impedance Zin = Z0 for all locations d on the line,
(2) Γ = 0, and (3) all of the incident power is delivered to
the load, regardless of the line length l. A summary of the
properties of standing waves is given in Table 2-4.
What is the difference
between the characteristic impedance Z0 and the input
impedance Zin ? When are they the same?
Concept Question 2-10:
A 50 Ω lossless transmission line is to be matched to a resistive
load impedance with ZL = 100 Ω via a quarter-wave section
as shown in Fig. 2-22, thereby eliminating reflections along
the feedline. Find the required characteristic impedance of the
quarter-wave transformer.
What is a quarter-wave transformer? How can it be used?
Concept Question 2-11:
A lossless transmission line
of length l is terminated in a short circuit. If l < λ /4, is
the input impedance inductive or capacitive?
Concept Question 2-12:
Feedline
A
λ/4 transformer
Z01 = 50 Ω
Z02
Zin
ZL = 100 Ω
Concept Question 2-13:
What is the input impedance
of an infinitely long line?
A'
λ/4
Figure 2-22 Configuration for Example 2-10.
Solution: To eliminate reflections at terminal AA′ , the input
impedance Zin looking into the quarter-wave line should be
equal to Z01 , which is the characteristic impedance of the
feedline. Thus, Zin = 50 Ω. From Eq. (2.97),
Zin =
2
Z02
,
ZL
If the input impedance of a
lossless line is inductive when terminated in a short
circuit, will it be inductive or capacitive when the line is
terminated in an open circuit?
Concept Question 2-14:
Exercise 2-14: A 50 Ω lossless transmission line uses
an insulating material with εr = 2.25. When terminated
in an open circuit, how long should the line be for its
input impedance to be equivalent to a 10 pF capacitor at
50 MHz?
Answer: l = 9.92 cm. (See
EM
.)
2-8 SPECIAL CASES OF THE LOSSLESS LINE
79
Table 2-4 Properties of standing waves on a lossless transmission line.
Voltage maximum
Voltage minimum
Positions of voltage maxima (also
positions of current minima)
Position of first maximum (also position of
first current minimum)
|Ve |max = |V0+ |[1 + |Γ|]
|Ve |min = |V + |[1 − |Γ|]
0
θr λ nλ
+
, n = 0, 1, 2, . . .
4π
2



 θr λ ,

if 0 ≤ θr ≤ π
4π
dmax =



 θr λ + λ , if − π ≤ θr ≤ 0
4π
2
dmax =
Position of first minimum (also position of
first current maximum)
θr λ (2n + 1)λ
+
,
4π
4
λ
θr
1+
dmin =
4
π
Input impedance
Zin = Z0
Positions at which Zin is real
at voltage maxima and minima
1 + |Γ|
Zin = Z0
1 − |Γ|
1 − |Γ|
Zin = Z0
1 + |Γ|
Positions of voltage minima (also positions
of current maxima)
Zin at voltage maxima
Zin at voltage minima
dmin =
zL + j tan β l
1 + jzL tan β l
Zin of short-circuited line
sc = jZ tan β l
Zin
0
Zin of open-circuited line
oc = − jZ cot β l
Zin
0
Zin of line of length l = nλ /2
Zin = ZL ,
Zin of line of length l = λ /4 + nλ /2
Zin = Z02 /ZL ,
Zin of matched line
Zin = Z0
n = 0, 1, 2, . . .
= Z0
1 + Γl
1 − Γl
n = 0, 1, 2, . . .
n = 0, 1, 2, . . .
|V0+ | = amplitude of incident wave; Γ = |Γ|e jθr with −π < θr < π ; θr in radians; Γl = Γe− j2β l .
Exercise 2-15: A 300 Ω feedline is to be connected to a
Answer: (a) Zin = 150 Ω, (b) S = 2, (c) Z0 = 212.1 Ω.
3-m long, 150 Ω line terminated in a 150 Ω resistor. Both
lines are lossless and use air as the insulating material,
and the operating frequency is 50 MHz. Determine (a) the
input impedance of the 3-m long line, (b) the voltage
standing-wave ratio on the feedline, and (c) the characteristic impedance of a quarter-wave transformer were it
to be used between the two lines in order to achieve S = 1
on the feedline. (See EM .)
Exercise 2-16: Through multiple trials, it was determined
that a load with unknown impedance ZL can be perfectly matched to a feedline with Zin = 50 Ω by using a
λ /4 transformer section with characteristic impedance of
60 Ω. What is ZL ?
Answer: ZL = 72 Ω.
80
CHAPTER 2
Technology Brief 3: Microwave Ovens
Percy Spencer, while working for Raytheon in the 1940s
on the design and construction of magnetrons for radar,
observed that a chocolate bar that had unintentionally
been exposed to microwaves had melted in his pocket.
The process of cooking by microwave was patented in
1946 and by the 1970s, microwave ovens had become
standard household items.
Microwave Absorption
A microwave is an electromagnetic wave whose frequency lies in the 300 MHz–300 GHz range (see
Fig. 1-16). When a material containing water is exposed
to microwaves, the water molecule reacts by rotating
itself in order to align its own electric dipole along the
direction of the oscillating electric field of the microwave.
The rapid vibration motion creates heat in the material,
resulting in the conversion of microwave energy into
thermal energy. The absorption coefficient of water, α ( f ),
exhibits a microwave spectrum that depends on the temperature of the water and the concentration of dissolved
TRANSMISSION LINES
salts and sugars present in it. If the frequency f is chosen
such that α ( f ) is high, the water-containing material
absorbs much of the microwave energy passing through
it and converts it to heat. However, this also means that
most of the energy is absorbed by a thin surface layer
of the material with not much energy remaining to heat
deeper layers. The penetration depth δp of a material,
defined as δp = 1/2α , is a measure of how deep the
power carried by an EM wave can penetrate into the
material. Approximately 95% of the microwave energy
incident upon a material is absorbed by the surface
layer of thickness 3δp . Figure TF3-1 displays calculated
spectra of δp for pure water and two materials with
different water contents.
◮ The frequency most commonly used in microwave
ovens is 2.45 GHz. The magnitude of δs at 2.45 GHz
varies beween ∼ 2 cm for pure water and 8 cm for a
material with a water content of only 20%. ◭
This is a practical range for cooking food in a microwave
oven; at much lower frequencies, the food is not a good
absorber of energy (in addition to the fact that the design
50
T = 20 ◦C
Penetration Depth δp (cm)
40
3δp
30
95% of energy
absorbed in
this layer
Chocolate bar
20
Food with 20% water
Microwave oven frequency (2.45 GHz)
10
Food
with 5
0% w
ater
Pure water
0
1
2
3
4
5
Frequency (GHz)
Figure TF3-1 Penetration depth as a function of frequency (1–5 GHz) for pure water and two foods with different water contents.
2-8 SPECIAL CASES OF THE LOSSLESS LINE
of the magnetron and the oven cavity become problematic), and at much higher frequencies, the microwave
energy cooks the food very unevenly (mostly the surface
layer). Whereas microwaves are readily absorbed by
water, fats, and sugars, they can penetrate through
most ceramics, glass, or plastics without loss of energy,
thereby imparting little or no heat to those materials.
Oven Operation
To generate high-power microwaves (∼ 700 watts), the
microwave oven uses a magnetron tube (Fig. TF3-2),
which requires the application of a voltage on the order
of 4000 volts. The typical household voltage of 115 volts
is increased to the required voltage level through a highvoltage transformer. The microwave energy generated
by the magnetron is transferred into a cooking chamber
designed to contain the microwaves within it through the
use of metal surfaces and safety Interlock switches.
81
◮ Microwaves are reflected by metal surfaces, so
they can bounce around the interior of the chamber,
be absorbed by the food, but not escape to the
outside. ◭
If the oven door is made of a glass panel, a metal
screen or a layer of conductive mesh is attached to it
to ensure the necessary shielding; microwaves cannot
pass through the metal screen if the mesh width is much
smaller than the wavelength of the microwave (λ ≈ 12 cm
at 2.5 GHz). In the chamber, the microwave energy
establishes a standing-wave pattern, which leads to an
uneven distribution. This is mitigated by using a rotating
metal stirrer that disperses the microwave energy to
different parts of the chamber.
Metal screen
Magnetron
Stirrer
Interlock switch
4,000 V
115 V
High-voltage transformer
Figure TF3-2 Microwave oven cavity.
82
CHAPTER 2
2-9 Power Flow on a Lossless
Transmission Line
Our discussion thus far has focused on the voltage and current
attributes of waves propagating on a transmission line. Now
we examine the flow of power carried by the incident and
reflected waves. We begin by reintroducing Eqs. (2.63a) and
(2.63b) with z = −d:
Ve (d) = V0+ (e jβ d + Γe− jβ d ),
V0+
˜ =
I(d)
Z0
(2.98a)
(e jβ d − Γe− jβ d ).
(2.98b)
In these expressions, the first terms represent the incident-wave
voltage and current, and the terms involving Γ represent the
reflected-wave voltage and current. The time-domain expressions for the voltage and current at location d from the load
are obtained by transforming Eq. (2.98) to the time domain:
υ (d,t) = Re[Ve e jω t ]
= Re[|V0+ |e
jφ +
Per our earlier discussion in connection with Eq. (1.31),
if the signs preceding ω t and β d in the argument of the
cosine term are both positive or both negative, then the cosine
term represents a wave traveling in the negative d direction.
Since d points from the load to the generator, the first term
in Eq. (2.100) represents the instantaneous incident power
traveling towards the load. This is the power that would be
delivered to the load in the absence of wave reflection (when
Γ = 0). Because β d is preceded by a minus sign in the
argument of the cosine of the second term in Eq. (2.100), that
term represents the instantaneous reflected power traveling in
the +d direction—away from the load. Accordingly, we label
these two power components
Pi (d,t) =
|V0+ |2
cos2 (ω t + β d + φ + ) (W),
Z0
Pr (d,t) = −|Γ|2
|V0+ |2
cos2 (ω t − β d + φ + + θr )
Z0
(2.101a)
(W).
(2.101b)
(e jβ d + |Γ|e jθr e− jβ d )e jω t ]
Using the trigonometric identity
= |V0+ |[cos(ω t + β d + φ + )
+ |Γ| cos(ω t − β d + φ + + θr )],
i(d,t) =
TRANSMISSION LINES
cos2 x = 21 (1 + cos2x),
(2.99a)
|V0+ |
[cos(ω t + β d + φ + )
Z0
the expressions in Eq. (2.101) can be rewritten as
− |Γ| cos(ω t − β d + φ + + θr )],
(2.99b)
Pi (d,t) =
jφ +
where we used the relations V0+ = |V0+ |e
and Γ = |Γ|e jθr
that were introduced earlier as Eqs. (2.31a) and (2.62), respectively.
2-9.1 Instantaneous Power
The instantaneous power carried by the transmission line is
equal to the product of υ (d,t) and i(d,t):
|V0+ |2
[1 + cos(2ω t + 2β d + 2φ + )],
2Z0
Pr (d,t) = −|Γ|2
(2.102a)
|V0+ |2
[1 + cos(2ω t − 2β d + 2φ + + 2θr )].
2Z0
(2.102b)
We note that in each case, the instantaneous power consists of
a dc term (not varying in time) and an ac term that oscillates at
an angular frequency of 2ω .
P(d,t) = υ (d,t) i(d,t)
◮ The power oscillates at twice the rate of the voltage or
current. ◭
= |V0+ |[cos(ω t + β d + φ + )
+ |Γ| cos(ω t − β d + φ + + θr )]
×
|V0+ |
[cos(ω t + β d + φ + )
Z0
2-9.2
− |Γ| cos(ω t − β d + φ + + θr )]
=
|V0+ |2
Z0
[cos2 (ω t + β d + φ + )
− |Γ|2 cos2 (ω t − β d + φ + + θr )]
(W).
(2.100)
Time-Average Power
From a practical standpoint, we usually are more interested
in the time-average power flowing along the transmission
line, Pav (d), than in the instantaneous power P(d,t). To compute Pav (d), we can use a time-domain approach or a computationally simpler phasor-domain approach. For completeness,
we consider both.
2-9 POWER FLOW ON A LOSSLESS TRANSMISSION LINE
83
Time-Domain Approach
The time-average power is equal to the instantaneous power
averaged over one time period T = 1/ f = 2π /ω . For the
incident wave, its time-average power is
i
Pav
(d) =
1
T
Z T
Pi (d,t) dt =
0
ω
2π
Z 2π /ω
i
Pav
=
2Z0
ZL
0
(W),
(2.104)
|V0+ |2
i
= −|Γ|2 Pav
.
2Z0
(2.105)
◮ The average reflected power is equal to the average
incident power, diminished by a multiplicative factor
of |Γ|2 . ◭
d=l
i
r
Note that the expressions for Pav
and Pav
are independent
of d, which means that the time-average powers carried by
the incident and reflected waves do not change as they travel
along the transmission line. This is as expected, because the
transmission line is lossless.
The net average power flowing towards (and then absorbed
by) the load shown in Fig. 2-23 is
|V0+ |2
[1 − |Γ|2]
2Z0
(W).
(2.106)
Phasor-Domain Approach
For any propagating wave with voltage and current phasors Ve
e a useful formula for computing the time-average power
and I,
is
h
i
Pav = 21 Re Ve · Ie∗ ,
(2.107)
d=0
Figure 2-23 The time-average power reflected by a load
connected to a lossless transmission line is equal to the incident
power multiplied by |Γ|2 .
e Application of this
where Ie∗ is the complex conjugate of I.
formula to Eqs. (2.98a) and (2.98b) gives
#
"
+∗
1
− j β d V0
∗ jβ d
+ jβ d
− jβ d
)
−Γ e )
(e
Pav = Re V0 (e + Γe
2
Z0
+2
|V0 |
1
(1 − |Γ|2 + Γe− j2β d − Γ∗ e j2β d )
= Re
2
Z0
+ 2 n
o
|V |
[1 − |Γ|2] + Re [|Γ|e− j(2β d−θr) − |Γ|e j(2β d−θr) ]
= 0
2Z0
=
i
r
Pav = Pav
+ Pav
=
r
i
Pav
= |Γ|2 Pav
−
Pi (d,t) dt. (2.103)
which is identical with the dc term of Pi (d,t) given by
Eq. (2.102a). A similar treatment for the reflected wave gives
r
Pav
= −|Γ|2
i
Pav
+
~
Vg
Upon inserting Eq. (2.102a) into Eq. (2.103) and performing
the integration, we obtain
|V0+ |2
Transmission line
Zg
=
|V0+ |2 [1 − |Γ|2] + |Γ|[cos(2β d − θr ) − cos(2β d − θr )]
2Z0
|V0+ |2
[1 − |Γ|2],
2Z0
(2.108)
which is identical to Eq. (2.106).
Exercise 2-17: For a 50 Ω lossless transmission line
terminated in a load impedance ZL = (100 + j50) Ω,
determine the fraction of the average incident power
reflected by the load.
Answer: 20%. (See
EM
.)
Exercise 2-18: For the line of Exercise 2-17, what is the
magnitude of the average reflected power if |V0+ | = 1 V?
r
Answer: Pav
= 2 (mW). (See
EM
.)
According to Eq. (2.102b),
the instantaneous value of the reflected power depends on
the phase of the reflection coefficient θr , but the average
reflected power given by Eq. (2.105) does not. Explain.
Concept Question 2-15:
84
CHAPTER 2
What is the average power
delivered by a lossless transmission line to a reactive load?
Concept Question 2-16:
What fraction of the incident
power is delivered to a matched load?
Concept Question 2-17:
Verify that
Z
2π t
1 T
1
+ β d + φ dt = ,
cos2
T 0
T
2
Concept Question 2-18:
regardless of the values of d and φ , so long as neither is a
function of t.
2-10 The Smith Chart
The Smith chart, developed by P. H. Smith in 1939, is a widely
used graphical tool for analyzing and designing transmissionline circuits. Even though it was originally intended to facilitate calculations involving complex impedances, the Smith
chart has become an important avenue for comparing and
characterizing the performance of microwave circuits. As the
material in this and the next section demonstrates, use of the
Smith chart not only avoids tedious manipulations of complex
numbers, but it also allows an engineer to design impedancematching circuits with relative ease.
2-10.1 Parametric Equations
The reflection coefficient Γ is, in general, a complex quantity
composed of a magnitude |Γ| and a phase angle θr or, equivalently, a real part Γr and an imaginary part Γi ,
Γ = |Γ|e jθr = Γr + jΓi ,
(2.109)
Similarly, point B represents ΓB = −0.5 − j0.2, or |ΓB | = 0.54
and θrB = 202◦ or, equivalently,
θrB = (360◦ − 202◦) = −158◦.
◮ When both Γr and Γi are negative, θr is in the
third quadrant in the Γr –Γi plane. Thus, when using
θr = tan−1 (Γi /Γr ) to compute θr , it may be necessary to
add or subtract 180◦ to obtain the correct value of θr . ◭
The unit circle shown in Fig. 2-24 corresponds to |Γ| = 1.
Because |Γ| ≤ 1 for a transmission line terminated with a
passive load, only that part of the Γr –Γi plane that lies within
the unit circle is useful to us; hence, future drawings will be
limited to the domain contained within the unit circle.
Impedances on a Smith chart are represented by their values
normalized to the line’s characteristic impedance, Z0 . From
Γ=
=
ZL /Z0 − 1
ZL /Z0 + 1
zL − 1
,
zL + 1
Γr = |Γ| cos θr ,
Γi = |Γ| sin θr .
(2.110a)
(2.110b)
The Smith chart lies in the complex Γ plane. In Fig. 2-24,
point A represents a reflection coefficient ΓA = 0.3 + j0.4 or,
equivalently,
zL =
1+Γ
.
1−Γ
(2.112)
The normalized load impedance zL is, in general, a complex
quantity composed of a normalized load resistance rL and a
normalized load reactance xL :
zL = rL + jxL .
(2.113)
Using Eqs. (2.109) and (2.113) in Eq. (2.112), we have
(1 + Γr ) + jΓi
,
(1 − Γr ) − jΓi
(2.114)
which can be manipulated to obtain explicit expressions for
each of rL and xL in terms of Γr and Γi . This is accomplished by
multiplying the numerator and denominator of the right-hand
side of Eq. (2.114) by the complex conjugate of the denominator and then separating the result into real and imaginary parts.
These steps lead to
|ΓA | = [(0.3)2 + (0.4)2]1/2 = 0.5
rL =
θrA = tan−1 (0.4/0.3) = 53◦ .
xL =
and
(2.111)
the inverse relation is
rL + jxL =
where
TRANSMISSION LINES
1 − Γ2r − Γ2i
,
(1 − Γr)2 + Γ2i
2Γi
.
(1 − Γr)2 + Γ2i
(2.115a)
(2.115b)
2-10 THE SMITH CHART
85
Γi
o
=
1
2
20
θrB
|Γ| = 1
θr = 180o
D
−1 −0.9 −0.7 −0.5 −0.3
Short-circuit
load
θr = 90o
B
|ΓB| = 0.54
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
−0.1
−0.2
−0.3
−0.4
−0.5
−0.6
−0.7
−0.8
−0.9
θrA = 53o
A
|ΓA| = 0.5
C
0.1
0.3
0.5
0.7
0.9
1
Γr
θr = 0o
Open-circuit
load
Unit circle
−1
θr = 270o or −90o
◦
◦
Figure 2-24 The complex Γ plane. Point A is at ΓA = 0.3 + j0.4 = 0.5e j53 , and point B is at ΓB = −0.5 − j0.2 = 0.54e j202 . The unit
circle corresponds to |Γ| = 1. At point C, Γ = 1, corresponding to an open-circuit load, and at point D, Γ = −1, corresponding to a short
circuit.
Equation (2.115a) implies that there exist many combinations
of values for Γr and Γi that yield the same value for the
normalized load resistance rL . For example, (Γr , Γi ) = (0.33, 0)
gives rL = 2, as does (Γr , Γi ) = (0.5, 0.29), as well as an
infinite number of other combinations. In fact, if we were to
plot in the Γr –Γi plane all possible combinations of Γr and Γi
corresponding to rL = 2, we would obtain the circle labeled
rL = 2 in Fig. 2-25. Similar circles can be obtained for other
values of rL . After some algebraic manipulations, Eq. (2.115a)
can be rearranged into the parametric equation for the circle
in the Γr –Γi plane corresponding to a given value of rL :
Γr −
rL
1 + rL
2
+ Γ2i =
1
1 + rL
2
.
(2.116)
The standard equation for a circle in the x–y plane with center
at (x0 , y0 ) and radius a is
(x − x0 )2 + (y − y0)2 = a2 .
(2.117)
Comparison of Eq. (2.116) with Eq. (2.117) shows that the
rL circle is centered at Γr = rL /(1 + rL ) and Γi = 0, and its
radius is 1/(1 + rL ). It therefore follows that all rL -circles
pass through the point (Γr , Γi ) = (1, 0). The largest circle
shown in Fig. 2-25 is for rL = 0, which also is the unit circle
corresponding to |Γ| = 1. This is to be expected, because when
rL = 0, |Γ| = 1 regardless of the magnitude of xL .
A similar manipulation of the expression for xL given by
Eq. (2.115b) leads to
2
1 2
1
(Γr − 1)2 + Γi −
=
,
xL
xL
(2.118)
86
CHAPTER 2
TRANSMISSION LINES
Γi
zL = rL + jxL
xL = 2
xL = 0.5
xL = 1
rL = 5
rL = ∞
xL = 0
rL = 0.5
rL = 0
xL = −0.5
Γr
xL = −1
rL = 2
rL = 1
xL = −2
Figure 2-25 Families of rL and xL circles within the domain |Γ| ≤ 1. The rL circles are concentric with the outermost circle corresponding
to rL = 0 and the innermost circle (of zero radius) corresponding to rL = ∞. The parts of the xL circles contained in the Smith chart are in
the upper half of the chart if xL is positive and in the lower half of the chart if xL is negative.
which is the equation of a circle of radius (1/xL ) centered
at (Γr , Γi ) = (1, 1/xL ). The xL circles in the Γr –Γi plane
are quite different from those for constant rL . To start with,
the normalized reactance xL may assume both positive and
negative values, whereas the normalized resistance cannot be
negative (negative resistances cannot be realized in passive
circuits). Hence, Eq. (2.118) represents two families of circles:
one for positive values of xL and another for negative ones.
Furthermore, as shown in Fig. 2-25, only part of a given circle
falls within the bounds of the |Γ| = 1 unit circle.
The families of circles of the two parametric equations given
by Eqs. (2.116) and (2.118) plotted for selected values of rL
and xL constitute the Smith chart shown in Fig. 2-26. The
Smith chart provides a graphical evaluation of Eqs. (2.115a
and b) and their inverses.
◮ Each point in the Smith chart conveys the values of two
interrelated quantities, namel zL and Γ. The intersection
of the rL and xL circles at that point defines zL = rL + jxL .
Simultaneously, the location of the point defines |Γ|
and θr . ◭
For example, point P in Fig. 2-26 represents a normalized
load impedance zL = 2 − j1, which corresponds to a voltage
reflection coefficient Γ = 0.45 exp(− j26.6◦ ). The magnitude
2-10 THE SMITH CHART
87
Outermost scale:
wavelengths toward
generator
0
0
R
Inner scale:
θr in degrees
0.25λ
0.25λ
O
|Г| = 0.45
P
Middle scale:
wavelengths
toward load
−26.6o
P
zL = 2 − j1
Figure 2-26 Point P represents a normalized load impedance zL = 2 − j1. The reflection coefficient has a magnitude |Γ| = OP/OR = 0.45
and an angle θr = −26.6◦ . Point R is an arbitrary point on the rL = 0 circle (which also is the |Γ| = 1 circle).
|Γ| = 0.45 is obtained by dividing the length of the line
between the center of the Smith chart and point P (designated
OP in Fig. 2-26), by OR, which is the length of the line
between the center of the Smith chart and the edge of the unit
circle (the radius of the unit circle corresponds to |Γ| = 1). The
perimeter of the Smith chart contains three concentric scales.
The innermost scale is labeled angle of reflection coefficient
in degrees. This is the scale for θr . As indicated in Fig. 2-26,
θr = −26.6◦ (−0.46 rad) for point P. The meanings and uses
of the other two scales are discussed next.
Exercise 2-19: Use the Smith chart to find the values of Γ
corresponding to the following normalized load impedances: (a) zL = 2 + j0, (b) zL = 1 − j1, (c) zL = 0.5 − j2,
(d) zL = − j3, (e) zL = 0 (short circuit), (f) zL = ∞ (open
circuit), and (g) zL = 1 (matched load).
Answer:
(a) Γ = 0.33, (b) Γ = 0.45 −63.4◦ ,
(c) Γ = 0.83 −50.9◦ , (d) Γ = 1 −36.9◦ , (e) Γ = −1,
(f) Γ = 1, (g) Γ = 0. (See EM .)
88
CHAPTER 2
TRANSMISSION LINES
SWR circle
|Г| = 0.45
Load
A
B
Input
0.287λ (location at A)
d
.1
=0
00
λ
B
d = 0.1λ
A
zL = 2 − j1
z(d)
0.387λ
(location at B)
Figure 2-27 Point A represents a normalized load zL = 2 − j1 at 0.287λ on the WTG scale. Point B represents the line input at d = 0.1λ
from the load. At B, z(d) = 0.6 − j0.66.
2-10.2 SWR Circle
Consider point A in the Smith chart of Fig. 2-27. At point A,
the normalized load impedance is zL = 2 − j1. The magnitude
of the corresponding reflection coefficient is
√
zL − 1
2
2 − j1 − 1
1 − j1
|Γ| =
=
=
= √ = 0.45.
zL + 1
2 − j1 + 1
3 − j1
10
Let us construct a circle centered at (Γr , Γi ) = (0, 0) and
passing through point A. Every point on this circle has the same
value for |Γ|, namely 0.45. This constant-|Γ| circle is also a
constant-SWR circle. This follows from the relation between
the voltage standing-wave ratio (SWR) and |Γ|, namely
S=
1 + |Γ|
.
1 − |Γ|
(2.119)
◮ A constant value of |Γ| corresponds to a constant value
of S, and vice versa. ◭
The utility of the SWR circle will become evident shortly.
2-10 THE SMITH CHART
89
2-10.3 Wave Impedance
From Eq. (2.74), the normalized wave impedance looking
toward the load at a distance d from the load is
Z(d) 1 + Γd
z(d) =
=
,
(2.120)
Z0
1 − Γd
where
(2.121)
Γd = Γe− j2β d = |Γ|e j(θr −2β d)
is the phase-shifted voltage reflection coefficient. The form of
Eq. (2.120) is identical with that for zL given by Eq. (2.112):
1+Γ
.
(2.122)
1−Γ
This similarity in form suggests that, if Γ is transformed into
Γd , zL gets transformed into z(d). On the Smith chart, the
transformation from Γ to Γd is achieved by maintaining |Γ|
constant and decreasing its phase θr by 2β d, which corresponds to a clockwise rotation (on the Smith chart) over an
angle of 2β d radians. A complete rotation around the Smith
chart corresponds to a phase change of 2π in Γ. The length d
corresponding to this phase change satisfies
zL =
2π
d = 2π ,
λ
from which it follows that d = λ /2.
2β d = 2
(2.123)
◮ The outermost scale around the perimeter of the
Smith chart (Fig. 2-26), called the wavelengths toward
generator (WTG) scale, has been constructed to denote
movement on the transmission line toward the generator,
in units of the wavelength λ . That is, d is measured in
wavelengths, and one complete rotation corresponds to
d = λ /2. ◭
In some transmission-line problems, it may be necessary to
move from some point on the transmission line toward a point
closer to the load, in which case the phase of Γ must be
increased, which corresponds to rotation in the counterclockwise direction. For convenience, the Smith chart contains a
third scale around its perimeter (in between the θr scale and
the WTG scale) for accommodating such an operation. It is
called the wavelengths toward load (WTL) scale.
Illustration:
Using the Smith Chart to Find Z(d)
To illustrate how the Smith chart is used to find Z(d),
consider a 50 Ω lossless transmission line terminated in
a load impedance ZL = (100 − j50) Ω. Our objective is to
find Z(d) at a distance d = 0.1λ from the load.
1. The normalized load impedance is
zL =
ZL 100 − j50
=
= 2 − j1
Z0
50
(point A in Fig. 2-27).
2. Point A is located at 0.287λ on the WTG scale.
3. Using a compass, we construct the SWR circle with its
center at the center of the Smith chart and radius passing
through A.
4. As was stated earlier, to transform zL to z(d), we need to
maintain |Γ| constant, which means staying on the SWR
circle, while decreasing the phase of Γ by 2β d radians.
This is equivalent to moving a distance d = 0.1λ toward
the generator on the WTG scale. Since point A is located
at 0.287λ on the WTG scale, z(d) is found by moving to
location 0.287λ + 0.1λ = 0.387λ on the WTG scale. A
radial line through this new position on the WTG scale
intersects the SWR circle at point B.
5. Point B represents z(d), whose value is
z(d) = 0.6 − j0.66.
To obtain Z(d), we unnormalize z(d) by multiplying it by
Z0 = 50 Ω:
Z(d) = (0.6 − j0.66) × 50 = (30 − j33) Ω.
This result can be verified analytically using Eq. (2.120).
The points between points A and B on the SWR circle
represent different locations along the transmission line.
If a line is of length l, its input impedance is Zin = Z0 z(l)
with z(l) determined by rotating a distance l from the load
along the WTG scale.
Exercise 2-20: Use the Smith chart to find the normalized
input impedance of a lossless line of length l terminated in
a normalized load impedance zL for each of the following
combinations: (a) l = 0.25λ , zL = 1 + j0; (b) l = 0.5λ ,
zL = 1 + j1; (c) l = 0.3λ , zL = 1 − j1; (d) l = 1.2λ , zL =
0.5 − j0.5; (e) l = 0.1λ , zL = 0 (short circuit); (f) l = 0.4λ ,
zL = j3; and (g) l = 0.2λ , zL = ∞ (open circuit).
(a) zin = 1 + j0, (b) zin = 1 + j1,
(c) zin = 0.76 + j0.84,
(d)
zin = 0.59 + j0.66,
(e) zin = 0 + j0.73, (f) zin = 0 + j0.72, (g) zin = 0 − j0.32.
(See EM .)
Answer:
90
CHAPTER 2
2-10.4 SWR, Voltage Maxima, and Minima
Consider a load with zL = 2 + j1. Figure 2-28 shows a Smith
chart with a SWR circle drawn through point A, representing zL . The SWR circle intersects the real (Γr ) axis at two
points, labeled Pmax and Pmin . At both points, Γi = 0 and
Γ = Γr . Also, on the real axis, the imaginary part of the load
impedance xL = 0. From the definition of Γ,
zL − 1
,
Γ=
zL + 1
(2.124)
It follows that points Pmax and Pmin correspond to
Γ = Γr =
r0 − 1
r0 + 1
(for Γi = 0),
A to Pmin . Since the location of |Ve |max corresponds to that of
e min and the location of |Ve |min corresponds to that of |I|
e max ,
|I|
the Smith chart provides a convenient way to determine the
distances from the load to all maxima and minima on the line
(recall that the standing-wave pattern has a repetition period of
λ /2).
2-10.5 Impedance to Admittance Transformations
In solving certain types of transmission line problems, it is
often more convenient to work with admittances than with
impedances. Any impedance Z is in general a complex quantity consisting of a resistance R and a reactance X:
(2.125)
where r0 is the value of rL where the SWR circle intersects
the Γr axis. Point Pmin corresponds to r0 < 1, and Pmax corresponds to r0 > 1. Rewriting Eq. (2.119) for |Γ| in terms of S,
we have
S−1
|Γ| =
.
(2.126)
S+1
Z = R + jX
Y=
1
R − jX
1
=
= 2
Z
R + jX
R + X2
The similarity in form of Eqs. (2.125) and (2.127) suggests
that S equals the value of the normalized resistance r0 . By
definition, S ≥ 1, and at point Pmax , r0 > 1, which further
satisfies the similarity condition. In Fig. 2-28, r0 = 2.6 at Pmax ;
hence S = 2.6.
◮ S is numerically equal to the value of r0 at Pmax , which
is the point at which the SWR circle intersects the real Γ
axis to the right of the chart’s center. ◭
Points Pmin and Pmax also represent locations on the line
where the magnitude of the voltage |Ve | is a minimum and
a maximum, respectively. This is easily demonstrated by
considering Eq. (2.121) for Γd . At point Pmax , the total phase
of Γd , that is, (θr − 2β d), equals zero or −2nπ (with n being
a positive integer), which is the condition corresponding to
|Ve |max , as indicated by Eq. (2.69). Similarly, at Pmin the total
phase of Γd equals −(2n + 1)π , which is the condition for
|Ve |min . Thus, for the transmission line represented by the SWR
circle shown in Fig. 2-28, the distance between the load and
the nearest voltage maximum is dmax = 0.037λ , which is
obtained by moving clockwise from the load at point A to
point Pmax , and the distance to the nearest voltage minimum is
dmin = 0.287λ , corresponding to the clockwise rotation from
(2.128)
(S).
(2.129)
The real part of Y is called the conductance G, and the
imaginary part of Y is called the susceptance B. That is,
Y = G + jB
(2.127)
(Ω).
The admittance Y is the reciprocal of Z:
For point Pmax , |Γ| = Γr ; hence
S−1
Γr =
.
S+1
TRANSMISSION LINES
(S).
(2.130)
Comparison of Eq. (2.130) with Eq. (2.129) reveals that
R
R2 + X 2
−X
B= 2
R + X2
G=
(S),
(2.131a)
(S).
(2.131b)
A normalized impedance z is defined as the ratio of Z to the
characteristic impedance of the line, Z0 . The same concept
applies to the definition of the normalized admittance y; that
is,
y=
Y
G
B
= + j = g + jb
Y0 Y0
Y0
(dimensionless), (2.132)
where Y0 = 1/Z0 is the characteristic admittance of the line
and
G
= GZ0
Y0
B
b=
= BZ0
Y0
g=
(dimensionless),
(2.133a)
(dimensionless).
(2.133b)
The lowercase quantities g and b represent the normalized
conductance and normalized susceptance of y, respectively.
Of course, the normalized admittance y is the reciprocal of the
normalized impedance z,
y=
Y
Z0
1
=
= .
Y0
Z
z
(2.134)
2-10 THE SMITH CHART
91
0.213λ
SW R
Distance to voltage
maximum from load
dmax = 0.037λ
A
Pmin
0
Pmax
0.25λ
r0 = 2.6
Pmin
Pmax A
zL = 2 + j1
Distance to voltage
minimum from load
dmax
dmin = 0.287λ
dmin
λ/4
Figure 2-28 Point A represents a normalized load with zL = 2 + j1. The standing-wave ratio is S = 2.6 (at Pmax ), the distance between the
load and the first voltage maximum is dmax = (0.25 − 0.213)λ = 0.037λ , and the distance between the load and the first voltage minimum
is dmin = (0.037 + 0.25)λ = 0.287λ .
Accordingly, using Eq. (2.122), the normalized load admittance yL is given by
2β d = 4π d/λ = 4πλ /4λ = π gives
z(d = λ /4) =
yL =
1−Γ
1
=
zL
1+Γ
(dimensionless).
(2.135)
Now let us consider the normalized wave impedance z(d) at
a distance d = λ /4 from the load. Using Eq. (2.120) with
1 + Γe− jπ
1−Γ
=
= yL .
1 − Γe− jπ
1+Γ
(2.136)
◮ Rotation by λ /4 on the SWR circle transforms z into y,
and vice versa. ◭
In Fig. 2-29, the points representing zL and yL are diametrically
opposite to each other on the SWR circle. In fact, such a
92
CHAPTER 2
TRANSMISSION LINES
Load impedance zL
A
B
Load admittance yL
Figure 2-29 Point A represents a normalized load zL = 0.6 + j1.4. Its corresponding normalized admittance is yL = 0.25 − j0.6, and it is
at point B.
transformation on the Smith chart can be used to determine
any normalized admittance from its corresponding normalized
impedance, and vice versa.
The Smith chart can be used with normalized impedances
or with normalized admittances. As an impedance chart, the
Smith chart consists of rL and xL circles: the resistance and
reactance of a normalized load impedance zL , respectively.
◮ When used as an admittance chart, the rL circles
become gL circles and the xL circles become bL circles,
where gL and bL are the conductance and susceptance of
the normalized load admittance yL , respectively. ◭
Example 2-11: Smith Chart Calculations
A 50 Ω lossless transmission line of length 3.3λ is terminated
by a load impedance ZL = (25 + j50) Ω. Use the Smith chart
to find (a) the voltage reflection coefficient, (b) the voltage
standing-wave ratio, (c) the distances of the first voltage
maximum and first voltage minimum from the load, (d) the
input impedance of the line, and (e) the input admittance of
the line.
Solution: (a) The normalized load impedance is
zL =
ZL
25 + j50
=
= 0.5 + j1,
Z0
50
2-10 THE SMITH CHART
93
0.135λ
θr = 83o
zL
O′
yin
A
E
|Г| = 0.62
S = 4.26
Location
~
of |V|min
C
0.25λ
B
O
Location
~
of |V|max
D
zin
l = 0.3λ
0.435λ
D
A
3.3λ
zin
zL = 0.5 + j1
dmax = 0.25λ − 0.135λ = 0.115λ
dmin = dmax + 0.25λ = 0.365λ
dmax
dmin
Figure 2-30 Solution for Example 2-11. Point A represents a normalized load zL = 0.5 + j1 at 0.135λ on the WTG scale. At A, θr = 83◦
and |Γ| = OA/OO′ = 0.62. At B, the standing-wave ratio is S = 4.26. The distance from A to B gives dmax = 0.115λ and from A to C
gives dmin = 0.365λ . Point D represents the normalized input impedance zin = 0.28 − j0.40, and point E represents the normalized input
admittance yin = 1.15 + j1.7.
which is marked as point A on the Smith chart in Fig. 2-30. A
radial line is drawn from the center of the chart at point O
through point A to the outer perimeter of the chart. The
line crosses the scale labeled “angle of reflection coefficient
in degrees” at θr = 83◦ . Next, measurements are made to
determine lengths OA and OO′ , of the lines between O and A
and between points O and O′ , respectively, where O′ is an
arbitrary point on the rL = 0 circle. The length OO′ is equal
to the radius of the |Γ| = 1 circle. The magnitude of Γ is then
94
CHAPTER 2
obtained from |Γ| = OA/OO′ = 0.62. Hence,
Γ = 0.62 83◦ .
the load, and that the distance between successive minima is
20 cm, find the load impedance.
(2.137)
(b) The SWR circle passing through point A crosses the Γr
axis at points B and C. The value of rL at point B is 4.26, from
which it follows that
S = 4.26.
(c) The first voltage maximum is at point B on the SWR
circle, which is at location 0.25λ on the WTG scale. The load,
represented by point A, is at 0.135λ on the WTG scale. Hence,
the distance between the load and the first voltage maximum
is
dmax = (0.25 − 0.135)λ = 0.115λ .
The first voltage minimum is at point C. Moving on the WTG
scale between points A and C gives
dmin = (0.5 − 0.135)λ = 0.365λ ,
which is 0.25λ past dmax .
(d) The line is 3.3λ long; subtracting multiples of 0.5λ leaves
0.3λ . From the load at 0.135λ on the WTG scale, the input
of the line is at (0.135 + 0.3)λ = 0.435λ . This is labeled as
point D on the SWR circle, and the normalized impedance is
zin = 0.28 − j0.40,
which yields
Solution: The distance between successive minima equals
λ /2. Hence, λ = 40 cm. In wavelength units, the first voltage
minimum is at
5
= 0.125λ .
dmin =
40
Point A on the Smith chart in Fig. 2-31 corresponds to S = 3.
Using a compass, the constant S circle is drawn through
point A. Point B corresponds to locations of voltage minima.
Upon moving 0.125λ from point B toward the load on the
WTL scale (counterclockwise), we arrive at point C, which
represents the location of the load. The normalized load
impedance at point C is
zL = 0.6 − j0.8.
Multiplying by Z0 = 50 Ω, we obtain
ZL = 50(0.6 − j0.8) = (30 − j40) Ω.
The outer perimeter of the
Smith chart represents what value of |Γ|? Which point on
the Smith chart represents a matched load?
Concept Question 2-19:
What is an SWR circle? What
quantities are constant for all points on an SWR circle?
Concept Question 2-20:
Zin = zin Z0 = (0.28 − j0.40)50 = (14 − j20) Ω.
(e) The normalized input admittance yin is found by moving
0.25λ on the Smith chart to the image point of zin across the
circle, labeled point E on the SWR circle. The coordinates of
point E give
yin = 1.15 + j1.7,
and the corresponding input admittance is
Yin = yinY0 =
TRANSMISSION LINES
yin
1.15 + j1.7
=
= (0.023 + j0.034) S.
Z0
50
Example 2-12:
Determining ZL Using
the Smith Chart
This problem is similar to Example 2-6, except that now we
demonstrate its solution using the Smith chart.
Given that the voltage standing-wave ratio S = 3 on a 50 Ω
line, that the first voltage minimum occurs at 5 cm from
What line length corresponds
to one complete rotation around the Smith chart? Why?
Concept Question 2-21:
Which points on the SWR circle correspond to locations of voltage maxima and minima
on the line and why?
Concept Question 2-22:
Given a normalized impedance zL , how do you use the Smith chart to find the
corresponding normalized admittance yL = 1/zL ?
Concept Question 2-23:
2-11 Impedance Matching
A transmission line usually connects a generator circuit at
one end to a load at the other. The load may be an antenna,
a computer terminal, or any circuit with an equivalent input
impedance ZL .
2-11 IMPEDANCE MATCHING
95
S = 3.0
Voltage min
A
B
C
Load
B
0.125λ
C
0.125λ
zL = ?
dmin
Figure 2-31 Solution for Example 2-12. Point A denotes that S = 3, point B represents the location of the voltage minimum, and point C
represents the load at 0.125λ on the WTL scale from point B. At C, zL = 0.6 − j0.8.
◮ The transmission line is said to be matched to the
load when its characteristic impedance Z0 = ZL , in which
case waves traveling on the line towards the load are not
reflected back to the source. ◭
Since the primary use of a transmission line is to transfer power
or transmit coded signals (such as digital data), a matched load
ensures that all of the power delivered to the transmission line
by the source is transferred to the load (and no echoes are
relayed back to the source).
The simplest solution to matching a load to a transmission
line is to design the load circuit such that its impedance
ZL = Z0 . Unfortunately, this may not be possible in practice
because the load circuit may have to satisfy other requirements. An alternative solution is to place an impedancematching network between the load and the transmission line
as illustrated in Fig. 2-32.
◮ The purpose of the matching network is to eliminate
reflections at terminals MM ′ for waves incident from
the source. Even though multiple reflections may occur
between AA′ and MM ′ , only a forward traveling wave
exists on the feedline. ◭
96
CHAPTER 2
TRANSMISSION LINES
Module 2.6 Interactive Smith Chart Locate the load on the Smith chart; display the corresponding reflection coefficient
and SWR circle; “move” to a new location at a distance d from the load; read the wave impedance Z(d) and phase-shifted
reflection coefficient Γd ; perform impedance to admittance transformations and vice versa; and use all of these tools to solve
transmission-line problems via the Smith chart.
Zg
~
Vg
M
Feedline
+
Z0
−
Generator
A
Matching
network
Zin
M'
ZL
A'
Load
Figure 2-32 The function of a matching network is to transform the load impedance ZL such that the input impedance Zin
looking into the network is equal to Z0 of the feedline.
Within the matching network, reflections can occur at both
terminals (AA′ and MM ′ ), creating a standing-wave pattern,
but the net result (of all of the multiple reflections within the
matching network) is that the wave incident from the source
experiences no reflection when it reaches terminals MM ′ . This
is achieved by designing the matching network to exhibit an
impedance equal to Z0 at MM ′ when looking into the network
from the transmission line side. If the network is lossless, then
all of the power going into it will end up in the load.
2-11 IMPEDANCE MATCHING
M
Feedline
Z0
97
λ/4 transformer
Zin
Z02
A
A'
Z0
d
λ/4
A
B
Z02
Feedline
Z01
B'
M'
Z0
A'
C
M'
d2
A'
ys(l1)
M
d1
A
Z0
Z0
M'
ZL
A'
Z0
y(d1)
M
Zin
ZL
ZL
(b) If ZL = complex: in-series λ/4 transformer inserted
at d = dmax or d = dmin
Feedline
Z0
L
(d) In-parallel insertion of inductor at distance d2
Z(d)
Z01 Zin
Zin
M'
(a) If ZL is real: in-series λ/4 transformer inserted at AA'
M
A
ZL
M'
Feedline
y(d2)
M
Feedline
A
Z0
d1
l1
ZL
A'
(c) In-parallel insertion of capacitor at distance d1
(e) In-parallel insertion of a short-circuited stub
Figure 2-33 Five examples of in-series and in-parallel matching networks.
◮ Matching networks may consist of lumped elements,
such as capacitors and inductors (but not resistors
because resistors incur ohmic losses), or of sections of
transmission lines with appropriate lengths and terminations. ◭
The matching network, which is intended to match a load
impedance ZL = RL + jXL to a lossless transmission line with
characteristic impedance Z0 , may be inserted either in series
(between the load and the feedline) as in Fig. 2-33(a) and (b)
or in parallel (Fig. 2-33(c) to (e)). In either case, the network
has to transform the real part of the load impedance from RL (at
the load) to Z0 at MM ′ in Fig. 2-32 and transform the reactive
part from XL (at the load) to zero at MM ′ . To achieve these two
transformations, the matching network must have at least two
degrees of freedom (that is, two adjustable parameters).
If XL = 0, the problem reduces to a single transformation, in which case matching can be realized by inserting
a quarter-wavelength transformer (Section 2-8.5) next to the
load (Fig. 2-33(a)).
◮ For the general case where XL 6= 0, a λ /4 transformer
can be designed to provide the desired matching, but it
has to be inserted at a distance dmax or dmin from the load
(Fig. 2-33(b)), where dmax and dmin are the distances to
voltage maxima and minima, respectively. ◭
The design procedure is outlined in Module 2.7. The
in-parallel insertion networks shown in Fig. 2-33(c)–(e) are the
subject of Examples 2-13 and 2-14.
2-11.1 Lumped-Element Matching
In the arrangement shown in Fig. 2-34, the matching network
consists of a single lumped element, either a capacitor or an
98
CHAPTER 2
Yd
M
Feedline
Y0
Yin
Ys
d
Feedline
Y0
M
Yin
YL
Yd
Ys
M'
M'
Shunt element
(a) Transmission-line circuit
TRANSMISSION LINES
Load
(b) Equivalent circuit
Figure 2-34 Inserting a reactive element with admittance Ys at MM ′ modifies Yd to Yin .
inductor, connected in parallel with the line at a distance d
from the load. Parallel connections call for working in the
admittance domain. Hence, the load is denoted by an admittance YL , and the line has a characteristic admittance Y0 .
The shunt element has an admittance Ys . At MM ′ , Yd is the
admittance due to the transmission-line segment to the right
of MM ′ . The input admittance Yin (referenced at a point just to
the left of MM ′ ) is equal to the sum of Yd and Ys :
Yin = Yd + Ys .
(2.138)
In general, Yd is complex, and Ys is purely imaginary because
it represents a reactive element (capacitor or inductor). Hence,
Eq. (2.138) can be written as
Yin = (Gd + jBd ) + jBs = Gd + j(Bd + Bs).
(2.139)
When all quantities are normalized to Y0 , Eq. (2.139) becomes
yin = gd + j(bd + bs).
(2.140)
To achieve a matched condition at MM ′ , it is necessary that
yin = 1 + j0, which translates into two specific conditions,
namely
gd = 1 (real-part condition),
(2.141a)
bs = −bd
(2.141b)
(imaginary-part condition).
The real-part condition is realized through the choice of d,
which is the distance from the load to the shunt element, and
the imaginary-part condition is realized through the choice of
lumped element (capacitor or inductor) and its value. These
two choices are the two degrees of freedom needed in order to
match the load to the feedline.
Example 2-13: Lumped Element
A load impedance ZL = 25 − j50 Ω is connected to a 50 Ω
transmission line. Insert a shunt element to eliminate reflections towards the sending end of the line. Specify the insert
location d (in wavelengths), the type of element, and its value
given that f = 100 MHz.
Solution: The normalized load impedance is
zL =
ZL
25 − j50
=
= 0.5 − j1,
Z0
50
which is represented by point A on the Smith chart of Fig. 2-35.
Next, we draw the constant S circle through point A. As alluded
to earlier, to perform the matching task, it is easier to work
with admittances than with impedances. The normalized load
admittance yL is represented by point B, which is obtained by
rotating point A over 0.25λ or equivalently by drawing a line
from point A through the chart center to the image of point A
on the S circle. The value of yL at B is
yL = 0.4 + j0.8,
and it is located at position 0.115λ on the WTG scale. In the
admittance domain, the rL circles become gL circles, and the
xL circles become bL circles. To achieve matching, we need to
move from the load toward the generator a distance d such that
the normalized input admittance yd of the line terminated in the
load (Fig. 2-34) has a real part of 1. This condition is satisfied
by either of the two matching points C and D on the Smith
charts of Figs. 2-35 and 2-36, respectively, corresponding to
intersections of the S circle with the gL = 1 circle. Points C
and D represent two possible solutions for the distance d in
Fig. 2-34(a).
2-11 IMPEDANCE MATCHING
99
0.115λ
0.063λ
Load yL
d1
First intersection of
gL = 1 circle with SWR circle.
At C, yd1 = 1 + j1.58.
B
C
gL = 1 circle
A
Yd1
d1 = 0.063λ
Feedline
Load zL
Y0
50 nH
YL = (0.4 + j0.8)Y0
First solution
Figure 2-35 Solution for point C of Example 2-13. Point A is the normalized load with zL = 0.5 − j1; point B is yL = 0.4 + j0.8. Point C
is the intersection of the SWR circle with the gL = 1 circle. The distance from B to C is d1 = 0.063λ .
Solution for Point C (Fig. 2-35): At C,
yd1 = 1 + j1.58,
Looking from the generator toward the parallel combination
of the line connected to the load and the shunt element, the
normalized input admittance at terminals MM ′ is
which is located at 0.178λ on the WTG scale. The distance
between points B and C is
yin1 = ys1 + yd1 ,
d1 = (0.178 − 0.115)λ = 0.063λ .
(continued on p. 101)
100
CHAPTER 2
TRANSMISSION LINES
Load yL
B
d2 = 0.207λ
D
A
Load zL
Second intersection of
gL = 1 circle with SWR circle.
At D, yd2 = 1 − j1.58.
Yd2
d2 = 0.207λ
Feedline
Y0
50 pF
YL = (0.4 + j0.8)Y0
Second solution
Figure 2-36 Solution for point D of Example 2-13. Point D is the second point of intersection of the SWR circle and the gL = 1 circle.
The distance B to D gives d2 = 0.207λ .
2-11 IMPEDANCE MATCHING
where ys1 is the normalized input admittance of the shunt
element. To match the feed line to the parallel combination, we
need yin1 = 1 + j0. Hence, we need ys1 to cancel the imaginary
part of yd1 ; that is,
ys1 = − j1.58.
101
Feedline
Y0
Yin
The corresponding impedance of the lumped element is
Zs1 =
d
M
Yd
M' Ys
Y0
Load
1
Z0
Z0
jZ0
1
=
=
=
=
= j31.62 Ω.
Ys1
ys1 Y0
jbs1
− j1.58 1.58
Y0
Since the value of Zs1 is positive, the element to be inserted
should be an inductor and its value should be
31.62
31.62
=
L=
= 50 nH.
ω
2π × 108
The results of this solution have been incorporated into the
circuit of Fig. 2-35.
YL
l
Shorted
stub
(a) Transmission line circuit
Feedline
M
Solution for Point D (Fig. 2-36): At point D,
yd2 = 1 − j1.58,
Yin
The needed normalized admittance of the reactive element is
Ys
M'
and the distance between points B and D is
d2 = (0.322 − 0.115)λ = 0.207λ .
Yd
(b) Equivalent circuit
Figure 2-37 Shorted-stub matching network.
ys2 = + j1.58.
Hence,
Zs2 = − j31.62 Ω,
which is the impedance of a capacitor with
C=
1
= 50 pF.
31.62ω
Figure 2-36 displays the circuit solution for d2 and C.
2-11.2 Single-Stub Matching
The single-stub matching network shown in Fig. 2-37(a)
consists of two transmission line sections: one of length d
connecting the load to the feedline at MM ′ and another of
length l connected in parallel with the other two lines at MM ′ .
This second line is called a stub, and it is usually terminated in
either a short or open circuit; hence, its input impedance and
admittance are purely reactive. The stub shown in Fig. 2-37(a)
has a short-circuit termination.
◮ The required two degrees of freedom are provided by
the length l of the stub and the distance d from the load to
the stub position. ◭
Because at MM ′ the stub is added in parallel to the line (which
is why it is called a shunt stub), it is easier to work with
admittances than with impedances. The matching procedure
consists of two steps. In the first step, the distance d is
selected to transform the load admittance, YL = 1/ZL , into an
admittance of the form Yd = Y0 + jB when looking toward
the load at MM ′ . Then in the second step, the length l of the
stub line is selected so that its input admittance Ys at MM ′ is
equal to − jB. The parallel sum of the two admittances at MM ′
yields Y0 , which is the characteristic admittance of the line.
The procedure is illustrated by Example 2-14.
Example 2-14: Single-Stub Matching
Repeat Example 2-13, but use a shorted stub (instead
of a lumped element) to match the load impedance
ZL = (25 − j50) Ω to the 50 Ω transmission line.
Solution: In Example 2-13, we demonstrated that the load
can be matched to the line via either of two solutions:
1. d1 = 0.063λ , and ys1 = jbs1 = − j1.58,
102
CHAPTER 2
2. d2 = 0.207λ , and ys2 = jbs2 = j1.58.
TRANSMISSION LINES
1. The normalized load impedance is
zL =
The locations of the insertion points at distances d1 and d2
remain the same, but now our task is to select corresponding
lengths l1 and l2 of shorted stubs that present the required
admittances at their inputs.
To determine l1 , we use the Smith chart in Fig. 2-38. The
normalized admittance of a short circuit is − j∞, which is
represented by point E on the Smith chart with position 0.25λ
on the WTG scale. A normalized input admittance of − j1.58
is located at point F with position 0.34λ on the WTG scale.
Hence,
which is shown as point A in the Smith chart of Fig. 2-40(b).
2. Move on the SWR circle until z(d) becomes purely real,
which occurs at points B and C.
(a) Point B:
d1 = (0.25 − 0.209)λ
= 0.041λ ,
z(d1 ) = 4.27,
l1 = (0.34 − 0.25)λ = 0.09λ .
Z(d1 ) = 4.27 × 50 = 213.5 Ω,
q
Z02 = Z(d1 ) Z01
√
= 213.5 × 50
Similarly, ys2 = j1.58 is represented by point G with position 0.16λ on the WTG scale of the Smith chart in Fig. 2-39.
Rotating from point E to point G involves a rotation of 0.25λ
plus an additional rotation of 0.16λ or
l2 = (0.25 + 0.16)λ = 0.41λ .
ZL
100 + j100
=
= 2 + j2,
Z01
50
= 103.3 Ω.
(b) Point C:
λ
4
= (0.041 + 0.25)λ = 0.291λ ,
d2 = d1 +
Example 2-15:
λ /4 Transformer for Complex
Load
Design a quarter-wavelength transformer to match a load with
ZL = (100 + j100) Ω to a 50 Ω line.
Solution: The required transmission-line circuit is shown in
Fig. 2-40(a) on p. 105. We need to insert a λ /4 transformer to
eliminate reflections at MM ′ . To that end, we need to specify
the insertion distance d and the characteristic impedance Z02
of the λ /4 section so that Zin = Z01 .
A λ /4 section with characteristic impedance Z02 transfers
an impedance Za at one end of the section to impedance Zb at
the other end such that
z(d2 ) = 0.23,
Z(d2 ) = 0.23 × 50 = 11.5 Ω,
q
Z02 = Z(d2 ) Z01
√
= 11.5 × 50
= 24.2 Ω.
To match an arbitrary load
impedance to a lossless transmission line through a matching network, what is the required minimum number of
degrees of freedom that the network should provide?
Concept Question 2-24:
In the case of the single-stub
matching network, what are the two degrees of freedom?
Concept Question 2-25:
Za Zb = Z022 .
(2.142)
In the present case (Fig. 2-40(a)), Za = Z(d), the input impedance at BB′ looking towards the load, and
Zb = Zin = Z01 = 50 Ω. Since the λ /4 transformer is a lossless
line, its impedance Z02 is purely real. Consequently, to satisfy
Eq. (2.142), it is necessary that Z(d) be purely real as well,
which is the key to how we select the distance d. Given this
rationale, we proceed as follows.
When a transmission line is
matched to a load through a single-stub matching network, no waves are reflected toward the generator. What
happens to the waves reflected by the load and by
the shorted stub when they arrive at terminals MM ′ in
Fig. 2-37?
Concept Question 2-26:
2-11 IMPEDANCE MATCHING
103
0.115λ
0.063λ
Load yL
d1
First intersection of
gL = 1 circle with SWR circle.
At C, yd1 = 1 + j1.58.
B
C
gL = 1 circle
Admittance of
short-circuit stub
(Example 2-14)
E
l1 = 0.09λ
A
F
Load zL
ys1 = −j1.58
Feedline
d1
YL = (0.4 + j0.8)Y0
l1
Figure 2-38 Solution for point C of Example 2-14. Point A is the normalized load with zL = 0.5 − j1; point B is yL = 0.4 + j0.8. Point C
is the intersection of the SWR circle with the gL = 1 circle. The distance from B to C is d1 = 0.063λ . The length of the shorted stub (E to F)
is l1 = 0.09λ .
104
CHAPTER 2
TRANSMISSION LINES
Load yL
ys2 = j1.58
G
B
d2 = 0.207λ
Admittance of
short circuit stub
(Example 2-14)
E
D
A
Second intersection of
gL = 1 circle with SWR circle.
At D, yd2 = 1 − j1.58.
l2 = 0.410λ
Load zL
Feedline
d2
YL = (0.4 + j0.8)Y0
l2
Figure 2-39 Solution for point D of Example 2-14. Point D is the second point of intersection of the SWR circle and the gL = 1 circle.
The distance B to D gives d2 = 0.207λ , and the distance E to G gives l2 = 0.410λ .
2-11 IMPEDANCE MATCHING
105
Z(d)
Feedline
M
d
λ/4
A
B
(a) Quarter-wave transformer
Z01 Zin
Z02
M'
Z01
B'
ZL
A'
zL = 2 + j2
0.209λ
SWR circle
d1
A
d2
C
B
(b) Smith chart solution
Figure 2-40 Solution for Example 2-15: (a) quarter-wave transformer and (b) Smith chart solution.
106
CHAPTER 2
TRANSMISSION LINES
Module 2.7 Quarter-Wavelength Transformer This module allows you to go through a multi-step procedure to design
a quarter-wavelength transmission line that, when inserted at the appropriate location on the original line, presents a matched
load to the feedline.
2-11 IMPEDANCE MATCHING
107
Module 2.8 Discrete Element Matching For each of two possible solutions, the module guides the user through a
procedure to match the feedline to the load by inserting a capacitor or an inductor at an appropriate location along the line.
108
CHAPTER 2
TRANSMISSION LINES
Module 2.9 Single-Stub Tuning Instead of inserting a lumped element to match the feedline to the load, this module
determines the length of a shorted stub that can accomplish the same goal.
2-12 Transients on Transmission Lines
Thus far, our treatment of wave propagation on transmission lines has focused on the analysis of single-frequency,
time-harmonic signals under steady-state conditions. The
impedance-matching and Smith chart techniques we developed, while useful for a wide range of applications, are
inappropriate for dealing with digital or wideband signals that
exist in digital chips, circuits, and computer networks. For such
signals, we need to examine the transient transmission line
response instead.
◮ The transient response of a voltage pulse on a transmission line is a time record of its back-and-forth travel
between the sending and receiving ends of the line, taking
into account all the multiple reflections (echoes) at both
ends. ◭
Let us start by considering the case of a single rectangular
pulse of amplitude V0 and duration τ , as shown in Fig. 2-41(a).
The amplitude of the pulse is zero prior to t = 0, V0 over
the interval 0 ≤ t ≤ τ , and zero afterwards. The pulse can
be described mathematically as the sum of two unit step
functions:
V (t) = V1 (t) + V2(t) = V0 u(t) − V0 u(t − τ ),
(2.143)
2-12 TRANSIENTS ON TRANSMISSION LINES
109
V(t)
V(t)
V0
V1(t) = V0 u(t)
V0
τ
τ
t
t
V2(t) = −V0 u(t − τ)
(a) Pulse of duration τ
(b) V(t) = V1(t) + V2(t)
Figure 2-41 A rectangular pulse V (t) of duration τ can be represented as the sum of two step functions of opposite polarities displaced
by τ relative to each other.
where the unit step function u(x) is
1 for x > 0,
u(x) =
0 for x < 0.
Rg
t=0
Transmission line
(2.144)
The first component, V1 (t) = V0 u(t), represents a dc voltage of amplitude V0 that is switched on at t = 0 and
then retains that value indefinitely. The second component,
V2 (t) = −V0 u(t − τ ), represents a dc voltage of amplitude −V0
that is switched on at t = τ and remains that way indefinitely.
As can be seen from Fig. 2-41(b), the sum V1 (t) + V2 (t) is
equal to V0 for 0 < t < τ and equal to zero for t < 0 and
t > τ . This representation of a pulse in terms of two step
functions allows us to analyze the transient behavior of the
pulse on a transmission line as the superposition of two dc
signals. Hence, if we can develop basic tools for describing
the transient behavior of a single step function, we can apply
the same tools for each of the two components of the pulse and
then add the results to obtain the response to V (t).
2-12.1 Transient Response to a Step Function
The circuit shown in Fig. 2-42(a) consists of a generator,
composed of a dc voltage source Vg and a series resistance Rg ,
connected to a lossless transmission line of length l and
characteristic impedance Z0 . The line is terminated in a purely
resistive load RL at z = l.
◮ Note that in previous sections z = 0 was defined as the
location of the load; now it is more convenient to define it
as the location of the source. ◭
The switch between the generator circuit and the transmission line is closed at t = 0. The instant the switch is closed,
+
Z0
Vg
RL
−
z
z=0
z=l
(a) Transmission-line circuit
Rg
+
Vg
−
I1+
+
V1+
Z0
−
(b) Equivalent circuit at t = 0+
Figure 2-42 At t = 0+ , immediately after closing the switch
in the circuit in part (a), the circuit can be represented by the
equivalent circuit in part (b). This is because from that instant
and until a reflection is received back from the load the generator
circuit “sees” only an impedance Z0 .
the transmission line appears to the generator circuit as a load
with impedance Z0 . This is because, in the absence of a signal
on the line, the input impedance of the line is unaffected by
the load impedance RL . The circuit representing the initial
condition is shown in Fig. 2-42(b). The initial current I1+
and corresponding initial voltage V1+ at the sending end of the
110
CHAPTER 2
V(z, T/2)
V(z, 3T/2)
(V1+ +
(V1+)
V
V1+
V(z, 5T/2)
(V1+ + V1− + V2+)
V1−)
(V1+)
V
V1+
V2+ = ΓGV1−
V1− = ΓLV1+
z
l/2
z
l
l/2
0
(a) V(z) at t = T/2
z
l
I(z, 5T/2)
I1− = −ΓL I1+
(I1+)
(I1+ + I1−)
(I1+)
I1+
I1+
I
I
(d) I(z) at t = T/2
(I1+ + I1− + I2+) (I + + I −)
1
1
I1+
I
z
l
l
(c) V(z) at t = 5T/2
I(z, 3T/2)
l/2
l/2
0
(b) V(z) at t = 3T/2
I(z, T/2)
0
(V1+ + V1−)
V
V1+
0
TRANSMISSION LINES
I2+ = −ΓG I1−
z
0
l/2
l
(e) I(z) at t = 3T/2
l/2
0
z
l
(f) I(z) at t = 5T/2
Figure 2-43 Voltage and current distributions on a lossless transmission line at t = T /2, t = 3T /2, and t = 5T /2 due to a unit step voltage
applied to a circuit with Rg = 4Z0 and RL = 2Z0 . The corresponding reflection coefficients are ΓL = 1/3 and Γg = 3/5.
transmission line are given by
I1+ =
Vg
,
Rg + Z0
V1+ = I1+ Z0 =
of V1+
mismatch generates a reflected wave with amplitude
I1+
Vg Z0
.
Rg + Z0
(2.145a)
(2.145b)
The combination
and
constitutes
√ a wave that travels
along the line with velocity up = 1/ µε immediately after
the switch is closed. The plus-sign superscript denotes the fact
that the wave is traveling in the +z direction. The transient
response of the wave is shown in Fig. 2-43 at each of three
instances in time for a circuit with Rg = 4Z0 and RL = 2Z0 . The
first response is at time t1 = T /2, where T = l/up is the time it
takes the wave to travel the full length of the line. By time t1 ,
the wave has traveled halfway down the line; consequently,
the voltage on the first half of the line is equal to V1+ , while the
voltage on the second half is still zero (Fig. 2-43(a)). At t = T ,
the wave reaches the load at z = l, and because RL 6= Z0 , the
V1− = ΓLV1+ ,
(2.146)
RL − Z0
RL + Z0
(2.147)
where
ΓL =
is the reflection coefficient of the load. For the specific case
illustrated in Fig. 2-43, RL = 2Z0 , which leads to ΓL = 1/3.
After this first reflection, the voltage on the line consists of
the sum of two waves: the initial wave V1+ and the reflected
wave V1− . The voltage on the transmission line at t2 = 3T /2 is
shown in Fig. 2-43(b); V (z, 3T /2) equals V1+ on the first half
of the line (0 ≤ z < l/2) and (V1+ + V1− ) on the second half
(l/2 ≤ z ≤ l).
At t = 2T , the reflected wave V1− arrives at the sending
end of the line. If Rg 6= Z0 , the mismatch at the sending end
generates a reflection at z = 0 in the form of a wave with
2-12 TRANSIENTS ON TRANSMISSION LINES
111
voltage amplitude V2+ given by
where x = ΓL Γg . The series inside the square bracket is the
geometric series of the function
V2+ = ΓgV1− = Γg ΓLV1+ ,
where
Γg =
(2.148)
Rg − Z0
Rg + Z0
(2.149)
is the reflection coefficient of the generator resistance Rg . For
Rg = 4Z0 , we have Γg = 0.6. As time progresses after t = 2T ,
the wave V2+ travels down the line toward the load and adds
to the previously established voltage on the line. Hence, at
t = 5T /2, the total voltage on the first half of the line is
V (z, 5T /2) = V1+ + V1− + V2+ = (1 + ΓL + ΓL Γg )V1+
(0 ≤ z < l/2),
V (z, 5T /2) = V1+ + V1− = (1 + ΓL )V1+
(l/2 ≤ z ≤ l).
(2.151)
The voltage distribution is shown in Fig. 2-43(c).
So far, we have examined the transient response of the
voltage wave V (z,t). The associated transient response of
the current I(z,t) is shown in Figs. 2-43(d)–(f). The current
behaves similarly to the voltage V (z,t), except for one important difference. Whereas at either end of the line the reflected
voltage is related to the incident voltage by the reflection
coefficient at that end, the reflected current is related to the
incident current by the negative of the reflection coefficient.
This property of wave reflection is expressed by Eq. (2.61).
Accordingly,
I2+
= −Γg I1−
= Γg ΓL I1+ ,
(2.152a)
(2.152b)
and so on.
◮ The multiple-reflection process continues indefinitely,
and the ultimate value that V (z,t) reaches as t approaches
+∞ is the same at all locations on the transmission line. ◭
The final value of V is given by
V∞ = V1+ +V1− +V2+ +V2− +V3+ +V3− +· · ·
= V1+ [1+ΓL +ΓL Γg +Γ2L Γg +Γ2L Γ2g +Γ3L Γ2g +· · · ]
= V1+ [(1+ΓL )(1+ΓL Γg +Γ2L Γ2g +· · · )]
= V1+ (1+ΓL )[1 + x + x2 + · · ·],
(2.153)
for |x| < 1.
(2.154)
Hence, Eq. (2.153) can be rewritten in the compact form
V∞ = V1+
1 + ΓL
.
1 − ΓL Γg
(2.155)
Upon replacing V1+ , ΓL , and Γg with Eqs. (2.145b), (2.147),
and (2.149) and simplifying the resulting expression, we obtain
V∞ =
(2.150)
while on the second half of the line the voltage is only
I1− = −ΓL I1+ ,
1
= 1 + x + x2 + · · ·
1−x
Vg RL
.
Rg + RL
(2.156)
The voltage V∞ is called the steady-state voltage on the line,
and its expression is exactly what we should expect on the
basis of dc analysis of the circuit in Fig. 2-43(a), where we
treat the transmission line as simply a connecting wire between
the generator circuit and the load. The corresponding steadystate current is
I∞ =
Vg
V∞
=
.
RL
Rg + RL
(2.157)
2-12.2 Bounce Diagrams
Keeping track of the voltage and current waves as they bounce
back and forth on the line is a rather tedious process. The
bounce diagram is a graphical presentation that allows us
to accomplish the same goal but with relative ease. The
horizontal axes in Figs. 2-44(a) and (b) represent the position
along the transmission line, while the vertical axes denote
time. Figures 2-44(a) and (b) pertain to V (z,t) and I(z,t),
respectively. The bounce diagram in Fig. 2-44(a) consists
of a zigzag line indicating the progress of the voltage wave
on the line. The incident wave V1+ starts at z = t = 0 and
travels in the +z direction until it reaches the load at z = l
at time t = T . At the very top of the bounce diagram, the
reflection coefficients are indicated by Γ = Γg at the generator
end and by Γ = ΓL at the load end. At the end of the first
straight-line segment of the zigzag line, a second line is drawn
to represent the reflected voltage wave V1− = ΓLV1+ . The
amplitude of each new straight-line segment equals the product
of the amplitude of the preceding straight-line segment and
the reflection coefficient at that end of the line. The bounce
diagram for the current I(z,t) in Fig. 2-44(b) adheres to the
same principle except for the reversal of the signs of ΓL and Γg
at the top of the bounce diagram.
112
CHAPTER 2
V
G = Gg
z=0
t=0
l/4
I
l/2
3l/4
V1+
G = GL
z=l
G = -Gg
l/4
z=0
t=0
T
GLV1+
4T
4T
Gg2 GL2 V1+
T
3T
Gg2 GL2 I1+
I(l/4, 4T)
5T
t
I1+
G = -GL
z=l
-Gg GL2 I1+
3T
V(l/4, 4T)
3l/4
Gg GL I1+
2T
Gg GL2 V1+
l/2
-GL I1+
Gg GLV1+
2T
TRANSMISSION LINES
t
5T
t
(a) Voltage bounce diagram
t
(b) Current bounce diagram
(1 + GL + Gg GL + Gg GL2 + Gg2 GL2)V1+
V(l/4, t)
GL = 1/3
Gg = 3/5
(1 +
GL)V1+
(1 + GL + Gg GL + Gg GL2)V1+
(1 + GL + Gg GL)V1+
V1+
V1+
t
T
4
T
7T 2T 9T
4
4
3T
15T 4T 17T
4
4
5T
(c) Voltage versus time at z = l/4
Figure 2-44 Bounce diagrams for (a) voltage and (b) current. In (c), the voltage variation with time at z = l/4 for a circuit with Γg = 3/5
and ΓL = 1/3 is deduced from the vertical dashed line at l/4 in (a).
Using the bounce diagram, the total voltage (or current) at
any point z1 and time t1 can be determined by first drawing
a vertical line through point z1 , and then adding the voltages
(or currents) of all the zigzag segments intersected by that
line between t = 0 and t = t1 . To find the voltage at z = l/4
and T = 4T , for example, we draw a dashed vertical line in
Fig. 2-44(a) through z = l/4, and we extend it from t = 0 to
t = 4T . The dashed line intersects four line segments. The total
voltage at z = l/4 and t = 4T therefore is
V (l/4, 4T ) = V1+ + ΓLV1+ + Γg ΓLV1+ + Γg Γ2LV1+
= V1+ (1 + ΓL + Γg ΓL + Γg Γ2L ).
The time variation of V (z,t) at a specific location z can be
obtained by plotting the values of V (z,t) along the (dashed)
2-12 TRANSIENTS ON TRANSMISSION LINES
113
vertical line passing through z. Figure 2-44(c) shows the
variation of V as a function of time at z = l/4 for a circuit
with Γg = 3/5 and ΓL = 1/3.
Γg = −0.6 ΓL = 0.5
+
5V
Example 2-16:
Pulse Propagation
(a) Pulse circuit
Γg = −0.6
l/4
z=0
t=0
1 ns
2 ns −4 V
Solution: The one-way propagation time is
3 ns
4 ns
5 ns
6 ns
7 ns
8 ns
9 ns
10 ns
11 ns
12 ns
0.6
l
= 2 ns.
=
c 3 × 108
The reflection coefficients at the load and the sending end are
RL − Z0 150 − 50
=
= 0.5,
RL + Z0 150 + 50
Rg − Z0
12.5 − 50
=
= −0.6.
Γg =
Rg + Z0
12.5 + 50
ΓL =
V1+
5 × 50
V01 Z0
=
= 4 V.
=
Rg + Z0 12.5 + 50
Using the information displayed in the bounce diagram, it
is straightforward to generate the voltage response shown in
Fig. 2-45(c). Note that at t = 2 ns, the voltage at the load
consists of both V1+ = 4 V, and its reflection V1− = ΓLV1+ = 2 V.
Thus, the total voltage at the load at t = 2 ns is 6 V. A
similar process occurs at 3 ns due to the negative step function,
which generates −6 V. Consequently, the sum of the two stepfunction contributions adds up to zero and stays that way until
t = 6 ns.
Example 2-17:
150 Ω
1 ns
The transmission-line circuit of Fig. 2-45(a) is excited by a
rectangular pulse of duration τ = 1 ns that starts at t = 0.
Establish the waveform of the voltage response at the load
given that the pulse amplitude is 5 V, the phase velocity is c,
and the length of the line is 0.6 m.
By Eq. (2.143), the pulse is treated as the sum of two step
functions: one that starts at t = 0 with an amplitude V10 = 5 V
and a second one that starts at t = 1 ns with an amplitude
V20 = −5 V. Except for the time delay of 1 ns and the sign
reversal of all voltage values, the two step functions generate
identical bounce diagrams, as shown in Fig. 2-45(b). For the
first step function, the initial voltage is given by
Z0 = 50 Ω
_
0
T=
12.5 Ω
t
ΓL = 0.5
l/2
3l/4
z=l
V1+ = 4 V
1 ns
2 ns
3 ns
2V
−2 V 4 ns
−1.2 V 5 ns
1.2 V
6 ns
7 ns
−0.6 V
0.6 V 8 ns
0.36 V 9 ns
10 ns
−0.36 V
11 ns
0.18 V
−0.18 V
12 ns
t
First step function
Second step function
(b) Bounce diagram
VL (V)
6V
4V
2V
−2 V
0.54 V
t (ns)
1 2 3 4 5 6 7 8 9 10 11 12
−1.8 V
−4 V
(c) Voltage waveform at the load
Figure 2-45 Example 2-16.
Time-Domain Reflectometer
A time-domain reflectometer (TDR) is an instrument used to
locate faults on a transmission line. Consider, for example, a
(formerly matched) long underground or undersea cable that
gets damaged at some distance d from the sending end of
the line. The damage may alter the electrical properties or the
shape of the cable, causing it to exhibit at the fault location
an effective resistance RLf . A TDR sends a step voltage down
114
CHAPTER 2
V(0, t)
Drop in level caused
by reflection from fault
V1+
6V
For a fault at a distance d, the round-trip time delay of the
echo is
2d
.
∆t =
up
From Fig. 2-46(a), ∆t = 12 µ s. Hence,
V1+ + V1−
3V
d=
t
0
12 μs
(a) Observed voltage at the sending end
TRANSMISSION LINES
∆t
12 × 10−6
up =
× 2.07 × 108 = 1, 242 m.
2
2
(c) The drop in level of V (0,t) shown in Fig. 2-46(a) represents V1− . Thus,
V1− = ΓfV1+ = −3 V,
or
t=0
−3
= −0.5,
6
where Γf is the reflection coefficient due to the effective fault
resistance RLf that appears at z = d.
From Eq. (2.59),
Γf =
Rg = Z0
Z0
+
Vg
−
z=0
Rf
Z0
RL = Z0
z=d
(b) The fault at z = d is represented by a
fault resistance Rf
Figure 2-46 Time-domain reflectometer of Example 2-17.
the line, and by observing the voltage at the sending end as a
function of time, it is possible to determine the location of the
fault and its severity.
If the voltage waveform shown in Fig. 2-46(a) is seen on
an oscilloscope connected to the input of a 75 Ω matched
transmission line, determine (a) the generator voltage, (b) the
location of the fault, and (c) the fault shunt resistance. The
line’s insulating material is Teflon with εr = 2.1.
Solution: (a) Since the line is properly matched, Rg =
RL = Z0 . In Fig. 2-46(b), the fault located a distance d from
the sending end is represented by a shunt resistance Rf . For a
matched line, Eq. (2.145b) gives
V1+ =
Vg Z0 Vg
Vg Z0
=
= .
Rg + Z0
2Z0
2
According to Fig. 2-46(a), V1+ = 6 V. Hence,
Vg = 2V1+ = 12 V.
(b) The propagation velocity on the line is
c
3 × 108
up = √ = √
= 2.07 × 108 m/s.
εr
2.1
Γf =
RLf − Z0
,
RLf + Z0
which leads to RLf = 25 Ω. This fault load is composed of the
parallel combination of the fault shunt resistance Rf and the
characteristic impedance Z0 of the line to the right of the fault:
1
1
1
= + ,
RLf
Rf Z0
so the shunt resistance must be 37.5 Ω.
Example 2-18: Pulse Transmission
When excited by a pulse source of amplitude 1 V and duration 5 µ s, a transmission line—with unknown characteristic
impedance Z0 and unknown load resistance RL —exhibits the
voltage response shown in Fig. 2-47(b) when observed at the
middle of the line. The line relative permittivity is εr = 2.25.
Determine (a) the length of the line, (b) Z0 , and (c) RL .
Solution: (a) The phase velocity is
c
3 × 108
= 2 × 108 m/s.
up = √ = √
εr
2.25
It takes 10 µ s for the initial part of the pulse to reach the middle
of the line, so it would take it T = 20 µ s to reach the end of
the line. Hence,
l = up T = 2 × 108 × 20 × 10−6 = 4000 m.
2-12 TRANSIENTS ON TRANSMISSION LINES
+
_
0
(c) The pulse that appears between 30 µ s and 35 µ s is due
to reflection by the load and its amplitude V1− = 0.3 V. Hence,
l
100 Ω
1V
115
Z0
5 μs
+
V(l/2)
_
ΓL =
RL
V1−
0.3
+ = 0.5 = 0.6.
V1
Also,
(a) Transmission line
ΓL =
RL − Z0
.
RL + Z0
Using ΓL = 0.6 and Z0 = 100 Ω leads to
V(l/2)
RL = 400 Ω.
0.5 V
0.3 V
Concept Question 2-27:
...
10 μs 15 μs
30 μs 35 μs
What is transient analysis
used for?
t
(b) Observed voltage at midpoint of the line
The transient analysis presented in this section was for a step voltage. How does
one use it for analyzing the response to a pulse?
Concept Question 2-28:
Figure 2-47 Circuit and voltage response of Example 2-18.
What is the difference
between the bounce diagram for voltage and the bounce
diagram for current?
Concept Question 2-29:
(b) From Eq. (2.145b),
V1+ =
Vg Z0
.
Rg + Z0
In the present case, Vg = 1 V, Rg = 100 Ω, and the waveform in
Fig. 2-47(b) indicates that V1+ = 0.5 V. Solving for Z0 leads to
Z0 = 100 Ω.
Concept Question 2-30:
What is a TDR and what is it
used for?
Concept Question 2-31:
What do V∞ and I∞ represent?
116
CHAPTER 2
TRANSMISSION LINES
Module 2.10 Transient Response For a lossless line terminated in a resistive load, the module simulates the dynamic
response—at any location on the line—to either a step or pulse waveform sent by the generator.
2-12 TRANSIENTS ON TRANSMISSION LINES
117
Technology Brief 4:
EM Cancer Zappers
From laser eye surgery to 3-D X-ray imaging, EM
sources and sensors have been used as medical diagnostic and treatment tools for many decades. Future
advances in information processing and other relevant
technologies will undoubtedly lead to the greater performance and utility of EM devices, as well as to the introduction of entirely new types of devices. This Technology
Brief introduces two recent EM technologies that are still
in their infancy but are quickly developing into serious
techniques for the surgical treatment of cancer tumors.
Microwave Ablation
Ultrasound transducer
Ablation catheter
(transmission line)
Liver
Ultrasound image
Figure TF4-1 Microwave ablation for liver cancer treatment.
In medicine, ablation is defined as the “surgical removal
of body tissue,” usually through the direct application of
chemical or thermal therapies.
◮ Microwave ablation applies the same heatconversion process used in a microwave oven (see
TB3), but instead of using microwave energy to
cook food, it is used instead to destroy cancerous
tumors by exposing them to a focused beam of
microwaves. ◭
The technique can be used percutaneously (through
the skin), laparoscopically (via an incision), or intraoperatively (open surgical access). Guided by an imaging
system, such as a CT scanner or an ultrasound imager,
the surgeon can localize the tumor and then insert a thin
coaxial transmission line (∼1.5 mm in diameter) directly
through the body to position the tip of the transmission
line (a probe-like antenna) inside the tumor (Fig. TF4-1).
The transmission line is connected to a generator capable of delivering 60 W of power at 915 MHz (Fig. TF4-2).
The rise in temperature of the tumor is related to the
amount of microwave energy it receives, which is equal
to the product of the generator’s power level and the
duration of the ablation treatment. Microwave ablation is
a promising new technique for the treatment of liver, lung,
and adrenal tumors.
High-Power Nanosecond Pulses
Bioelectrics is an emerging field focused on the study
of how electric fields behave in biological systems. Of
particular recent interest is the desire to understand
Figure TF4-2 Photograph of the setup for a percutaneous
microwave ablation procedure in which three single microwave
applicators are connected to three microwave generators. (Courtesy of RadioGraphics, October 2005, pp. 569–583.)
how living cells might respond to the application of
extremely short pulses (on the order of nanoseconds
(10−9 s) and even as short as picoseconds (10−12 s)) with
exceptionally high voltage and current amplitudes.
118
CHAPTER 2
TRANSMISSION LINES
3 When the
trailing edges
of the waves
finally meet, the
pulse ends.
1 With the switch open, the
device is charged up by its
connection to the high-voltage
source. Closing the switch sets
up transient waves.
2
The voltage waves
reflect off the ends of the
transmission line. The wave near
the switch inverts (red)—its polarity
changes—when it reflects, because that end is
shorted. When the inverted and noninverted waves crash
into each other at the load, a pulse of voltage results.
Figure TF4-3 High-voltage nanosecond pulse delivered to tumor cells via a transmission line. The cells to be shocked by the pulse sit in
a break in one of the transmission-line conductors. (Courtesy of IEEE Spectrum, August 2006.)
◮ The motivation is to treat cancerous cells by
zapping them with high-power pulses. The pulse
power is delivered to the cell via a transmission line,
as illustrated by the example in Fig. TF4-3. ◭
Note that the pulse is about 200 ns long, and its voltage
and current amplitudes are approximately 3,000 V and
60 A, respectively. Thus, the peak power level is about
180,000 W! However, the total energy carried by the
pulse is only (1.8 × 105 ) × (2 × 10−7) = 0.0036 joules.
Despite the low energy content, the very high voltage
appears to be very effective in destroying malignant
tumors (in mice, so far), with no regrowth.
IMPORTANT TERMS
119
Chapter 2 Summary
Concepts
• A transmission line is a two-port network connecting a
generator to a load. EM waves traveling on the line may
experience ohmic power losses, dispersive effects, and
reflections at the generator and load ends of the line.
These transmission-line effects may be ignored if the
line length is much shorter than λ .
• TEM transmission lines consist of two conductors that
can support the propagation of transverse electromagnetic waves characterized by electric and magnetic
fields that are transverse to the direction of propagation.
TEM lines may be represented by a lumped-element
model consisting of four line parameters (R′ , L′ , G′ ,
and C ′ ) whose values are specified by the specific line
geometry, the constitutive parameters of the conductors
and of the insulating material between them, and the
angular frequency ω .
• Wave propagation on a transmission line, which is
represented by the phasor voltage Ve (z) and associated
e
current I(z),
is governed by the propagation constant
of the line, γ = α + jβ , and its characteristic impedance Z0 . Both γ and Z0 are specified by ω and the four
line parameters.
• If R′ = G′ = 0, the line becomes lossless (α = 0). A
lossless line is generally nondispersive, meaning that
the phase velocity of a wave is independent of the
frequency.
• In general, a line supports two waves, an incident wave
supplied by the generator and another wave reflected
•
•
•
•
•
by the load. The sum of the two waves generates a
standing-wave pattern with a period of λ /2. The voltage standing-wave ratio S, which is equal to the ratio
of the maximum to minimum voltage magnitude on the
line, varies between 1 for a matched load (ZL = Z0 ) to ∞
for a line terminated in an open circuit, a short circuit,
or a purely reactive load.
The input impedance of a line terminated in a short circuit or open circuit is purely reactive. This property can
be used to design equivalent inductors and capacitors.
The fraction of the incident power delivered to the load
by a lossless line is equal to (1 − |Γ|2).
The Smith chart is a useful graphical tool for analyzing transmission-line problems and for designing
impedance-matching networks.
Matching networks are placed between the load and
the feed transmission line for the purpose of eliminating reflections toward the generator. A matching
network may consist of lumped elements in the form
of capacitors and/or inductors, or it may consist of
sections of transmission lines with appropriate lengths
and terminations.
Transient analysis of pulses on transmission lines can
be performed using a bounce-diagram graphical technique that tracks reflections at both the load and generator ends of the transmission line.
120
CHAPTER 2
TRANSMISSION LINES
Mathematical and Physical Models
TEM Transmission Lines
L′C ′ = µε
′
G
σ
=
C′
ε
p
(R′ + jω L′ )(G′ + jω C ′ )
(Np/m)
p
β = Im(γ ) = Im
(R′ + jω L′ )(G′ + jω C ′ )
(rad/m)
s
R′ + jω L′
R′ + jω L′
=
(Ω)
Z0 =
γ
G′ + j ω C ′
α = Re(γ ) = Re
Γ=
zL − 1
zL + 1
Step Function Transient Response
Vg Z0
Rg + Z0
Vg RL
V∞ =
Rg + RL
Rg − Z0
Γg =
Rg + Z0
RL − Z0
ΓL =
RL + Z0
Lossless Line
α =0
√
β = ω L′C ′
r
L′
Z0 =
C′
1
up = √
µε
λ=
up
λ0
c 1
= √ =√
f
f εr
εr
dmax =
dmin =
V1+ =
S=
(m/s)
θr λ nλ
+
4π
2
θr λ (2n + 1)λ
+
4π
4
1 + |Γ|
1 − |Γ|
Pav =
|V0+ |2
[1 − |Γ|2]
2Z0
PROBLEMS
Important Terms
121
Provide definitions or explain the meaning of the following terms:
admittance Y
air line
attenuation constant α
bounce diagram
characteristic impedance Z0
coaxial line
complex propagation constant γ
conductance G
current maxima and minima
dispersive transmission line
distortionless line
effective relative permittivity εeff
guide wavelength λ
higher-order transmission lines
impedance matching
in-phase
input impedance Zin
load impedance ZL
lossless line
lumped-element model
standing wave
standing-wave pattern
surface resistance Rs
susceptance B
SWR circle
telegrapher’s equations
TEM transmission lines
time-average power Pav
transient response
transmission-line parameters
two-wire line
unit circle
voltage maxima and minima
voltage reflection coefficient Γ
voltage standing-wave ratio
(VSWR or SWR) S
wave equations
wave impedance Z(d)
waveguide
WTG and WTL
matched transmission line
matching network
microstrip line
normalized impedance
normalized load reactance xL
normalized load resistance rL
open-circuited line
optical fiber
parallel-plate line
perfect conductor
perfect dielectric
phase constant β
phase opposition
phase-shifted reflection
coefficient Γd
quarter-wave transformer
short-circuited line
single-stub matching
slotted line
Smith chart
PROBLEMS
i(z, t)
Sections 2-1 to 2-4: Transmission-Line Model
R' ∆z
2
L' ∆z
2
R' ∆z
2
+
2.1 A transmission line of length l connects a load to a
sinusoidal voltage source with an oscillation frequency f .
Assuming that the velocity of wave propagation on the line
is c, for which of the following situations is it reasonable to
ignore the presence of the transmission line in the solution of
the circuit?
∗ (a) l = 20 cm, f = 20 kHz
(b) l = 50 km, f = 60 Hz
∗ (c) l = 20 cm, f = 600 MHz
υ(z, t)
G' ∆z
L' ∆z
2 i(z + ∆z, t)
+
C' ∆z
υ(z + ∆z, t)
−
−
∆z
Figure P2.3 Transmission-line model for Problem 2.3.
(d) l = 1 mm, f = 100 GHz
2.2 A two-wire copper transmission line is embedded in a
dielectric material with εr = 2.6 and σ = 2 × 10−6 S/m. Its
wires are separated by 3 cm and their radii are 1 mm each.
(a) Calculate the line parameters R′ , L′ , G′ , and C′ at 2 GHz.
(b) Compare your results with those based on Module 2.1.
Include a printout of the screen display.
2.3 Show that the transmission-line model shown in
Fig. P2.3 yields the same telegrapher’s equations given by
Eqs. (2.14) and (2.16).
∗ Answer(s) available in Appendix E.
∗ 2.4 A 1 GHz parallel-plate transmission line consists of
1.2-cm wide copper strips separated by a 0.15-cm thick
layer of polystyrene. Appendix B gives µc = µ0 = 4π × 10−7
(H/m) and σc = 5.8 × 107 (S/m) for copper, and εr = 2.6 for
polystyrene. Use Table 2-1 to determine the line parameters
of the transmission line. Assume that µ = µ0 and σ ≈ 0 for
polystyrene.
2.5 For the parallel-plate transmission line of Problem 2.4,
the line parameters are given by R′ = 1 Ω/m, L′ = 167 nH/m,
G′ = 0, and C′ = 172 pF/m. Find α , β , up , and Z0 at 1 GHz.
122
CHAPTER 2
2.6 A coaxial line with inner and outer conductor diameters
of 0.5 cm and 1 cm, respectively, is filled with an insulating
material with εr = 4.5 and σ = 10−3 S/m. The conductors are
made of copper.
(a) Calculate the line parameters at 1 GHz.
(b) Compare your results with those based on Module 2.2.
Include a printout of the screen display.
2.7 Find α , β , up , and Z0 for the coaxial line of Problem 2.6.
Verify your results by applying Module 2.2. Include a printout
of the screen display.
∗ 2.8 Find α , β , u , and Z for the two-wire line of Problem
p
0
2.2. Compare results with those based on Module 2.1. Include
a printout of the screen display.
previously mentioned features of the lossless line. Show that
for a distortionless line,
α =R
′
r
C′ √ ′ ′
= RG ,
L′
√
β = ω L′C ′ ,
r
L′
Z0 =
.
C′
∗ 2.14 For a distortionless line (see Problem 2.13) with
Z0 = 50 Ω, α = 20 (mNp/m), and up = 2.5 × 108 (m/s), find
the line parameters and λ at 100 MHz.
2.15 Find α and Z0 of a distortionless line whose R′ = 2 Ω/m
and G′ = 2 × 10−4 S/m.
Section 2-5: Microstrip
2.9 A lossless microstrip line uses a 1-mm wide conducting
strip over a 1-cm thick substrate with εr = 2.5. Determine the
line parameters, εeff , Z0 , and β at 10 GHz. Compare your
results with those obtained by using Module 2.3. Include a
printout of the screen display.
∗ 2.10 Use Module 2.3 to design a 100 Ω microstrip transmis-
sion line. The substrate thickness is 1.8 mm and its εr = 2.3.
Select the strip width w, and determine the guide wavelength λ
at f = 5 GHz. Include a printout of the screen display.
2.11 A 50 Ω microstrip line uses a 0.6 mm alumina substrate
with εr = 9. Use Module 2.3 to determine the required strip
width w. Include a printout of the screen display.
2.12 Generate a plot of Z0 as a function of strip width w
over the range from 0.05 mm to 5 mm for a microstrip line
fabricated on a 0.7-mm thick substrate with εr = 9.8.
Section 2-6: Lossless Line
∗ 2.16 A transmission line operating at 125 MHz has
Z0 = 40 Ω, α = 0.02 (Np/m), and β = 0.75 rad/m. Find the
line parameters R′ , L′ , G′ , and C ′ .
2.17 Using a slotted line, the voltage on a lossless transmission line was found to have a maximum magnitude of 1.5 V
and a minimum magnitude of 0.6 V. Find the magnitude of the
load’s reflection coefficient.
∗ 2.18 Polyethylene with ε = 2.25 is used as the insulating
r
material in a lossless coaxial line with a characteristic impedance of 50 Ω. The radius of the inner conductor is 1.2 mm.
(a) What is the radius of the outer conductor?
(b) What is the phase velocity of the line?
2.19 A 50 Ω lossless transmission line is terminated in a load
with impedance ZL = (30 − j50) Ω. The wavelength is 8 cm.
Determine:
(a) The reflection coefficient at the load.
2.13 In addition to not dissipating power, a lossless line
has two important features: (1) it is dispersionless (up is
independent of frequency), and (2) its characteristic impedance Z0 is purely real. Sometimes, it is not possible to design
a transmission line such that R′ ≪ ω L′ and G′ ≪ ω C ′ , but it is
possible to choose the dimensions of the line and its material
properties to satisfy the condition
R′C ′ = L′ G′
TRANSMISSION LINES
(distortionless line)
Such a line is called a distortionless line, because despite
the fact that it is not lossless, it nonetheless possesses the
(b) The standing-wave ratio on the line.
(c) The position of the voltage maximum nearest the load.
(d) The position of the current maximum nearest the load.
(e) Verify quantities in parts (a)–(d) using Module 2.4. Include a printout of the screen display.
2.20 A 300 Ω lossless air transmission line is connected to a
complex load composed of a resistor in series with an inductor,
as shown in Fig. P2.20. At 5 MHz, determine: (a) Γ, (b) S,
(c) location of voltage maximum nearest to the load, and (d)
location of current maximum nearest to the load.
PROBLEMS
123
R = 600 Ω
RL = 75 Ω
Z0 = 50 Ω
Z0 = 300 Ω
L = 0.02 mH
C=?
Figure P2.20 Circuit for Problem 2.20.
Figure P2.26 Circuit for Problem 2.26.
Section 2-7: Wave and Input Impedance
∗ 2.21 On a 150 Ω lossless transmission line, the following
observations were noted: distance of first voltage minimum
from the load = 3 cm; distance of first voltage maximum from
the load = 9 cm; and S = 3. Find ZL .
2.22 Using a slotted line, the following results were
obtained: distance of first minimum from the load = 4 cm;
distance of second minimum from the load = 14 cm; and
voltage standing-wave ratio = 1.5. If the line is lossless and
Z0 = 50 Ω, find the load impedance.
∗ 2.27 At an operating frequency of 300 MHz, a lossless 50 Ω
air-spaced transmission line 2.5 m in length is terminated with
an impedance ZL = (40 + j20) Ω. Find the input impedance.
2.28 A lossless transmission line of electrical length
l = 0.35λ is terminated in a load impedance as shown in
Fig. P2.28. Find Γ, S, and Zin . Verify your results using
Modules 2.4 or 2.5. Include a printout of the screen’s output
display.
l = 0.35λ
∗ 2.23 A load with impedance Z = (25 − j50) Ω is to be
L
connected to a lossless transmission line with the characteristic
impedance Z0 . Z0 is chosen such that the standing-wave ratio
is the smallest possible. What should Z0 be?
2.24 A 50 Ω lossless line terminated in a purely resistive load
has a voltage standing-wave ratio of 3. Find all possible values
of ZL .
2.25 Apply Module 2.4 to generate plots of the voltage
standing-wave pattern for a 50 Ω line terminated in a load
impedance ZL = (100 − j50) Ω. Set Vg = 1 V, Zg = 50 Ω,
εr = 2.25, l = 40 cm, and f = 1 GHz. Also determine S, dmax ,
and dmin .
2.26 A 50 Ω lossless transmission line is connected to a
load composed of a 75 Ω resistor in series with a capacitor
of unknown capacitance (Fig. P2.26). If at 10 MHz the
voltage standing-wave ratio on the line was measured to be 3,
determine the capacitance C.
Zin
Z0 = 100 Ω
ZL = (60 + j30) Ω
Figure P2.28 Circuit for Problem 2.28.
2.29 Show that the input impedance of a quarter-wavelength
lossless line terminated in a short circuit appears as an open
circuit.
2.30 Show that, at the position where the magnitude of the
voltage on the line is a maximum, the input impedance is
purely real.
2.31 A voltage generator with
υg (t) = 5 cos(2π × 109t) V
and internal impedance Zg = 50 Ω is connected to a 50 Ω
lossless air-spaced transmission line. The line length is
5 cm and the line is terminated in a load with impedance
ZL = (100 − j100) Ω. Determine:
124
CHAPTER 2
∗ (a) Γ at the load.
TRANSMISSION LINES
All lines are 50 Ω and lossless.
∗ (a) Calculate Z , the input impedance of the antennain1
(b) Zin at the input to the transmission line.
ei and input current I˜i .
(c) The input voltage V
(d) The quantities in parts (a)–(c) using Modules 2.4 or 2.5.
2.32 A 6 m section of 150 Ω lossless line is driven by a
source with
vg (t) = 5 cos(8π × 107t − 30◦)
(V)
and Zg = 150 Ω. If the line, which has a relative permittivity
εr = 2.25, is terminated in a load ZL = (150 − j50) Ω, determine:
(a) λ on the line,
terminated line, at the parallel juncture.
(b) Combine Zin1 and Zin2 in parallel to obtain ZL′ , the effective load impedance of the feedline.
(c) Calculate Zin of the feedline.
2.34 A 50 Ω lossless line is terminated in a load impedance
ZL = (30 − j20) Ω.
(a) Calculate Γ and S.
(b) It has been proposed that by placing an appropriately
selected resistor across the line at a distance dmax from
the load (as shown in Fig. P2.34(b)), where dmax is the
distance from the load of a voltage maximum, then it is
possible to render Zi = Z0 , thereby eliminating reflection
back to the end. Show that the proposed approach is valid
and find the value of the shunt resistance.
∗ (b) the reflection coefficient at the load,
(c) the input impedance,
(d) the input voltage Vei ,
Z0 = 50 Ω
ZL = (30 − j20) Ω
(e) the time-domain input voltage vi (t), and
(f) quantities in parts (a)–(d) using Modules 2.4 or 2.5.
(a)
2.33 Two half-wave dipole antennas, each with an impedance of 75 Ω, are connected in parallel through a pair of
transmission lines, and the combination is connected to a feed
transmission line, as shown in Fig. P2.33.
dmax
Z0 = 50 Ω
λ
0.2
0.3λ
Zin
ZL = (30 − j20) Ω
R
Zi
75 Ω
(Antenna)
Zin1
Zin2
(b)
Figure P2.34 Circuit for Problem 2.34.
0.2
λ
75 Ω
(Antenna)
Figure P2.33 Circuit for Problem 2.33.
∗ 2.35 For the lossless transmission line circuit shown in
Fig. P2.35, determine the equivalent series lumped-element
circuit at 400 MHz at the input to the line. The line has a
characteristic impedance of 50 Ω and the insulating layer has
εr = 2.25.
PROBLEMS
Zin
125
Z0 = 50 Ω
75 Ω
1.2 m
2.41 A 50 Ω lossless line of length l = 0.375λ connects a 300
MHz generator with Veg = 300 V and Zg = 50 Ω to a load ZL .
Determine the time-domain current through the load for:
(a) ZL = (50 − j50) Ω
∗ (b) Z = 50 Ω
L
(c) ZL = 0 (short circuit)
For part (a), verify your results by deducing the information
you need from the output products generated by Module 2.4.
Figure P2.35 Circuit for Problem 2.35.
Section 2-9: Power Flow on Lossless Line
Section 2-8: Special Cases of the Lossless Line
2.36 At an operating frequency of 300 MHz, it is desired to
use a section of a lossless 50 Ω transmission line terminated
in a short circuit to construct an equivalent load with reactance
X = 40 Ω. If the phase velocity of the line is 0.75c, what is
the shortest possible line length that would exhibit the desired
reactance at its input? Verify your result using Module 2.5.
∗ 2.37 A lossless transmission line is terminated in a short
circuit. How long (in wavelengths) should the line be for it
to appear as an open circuit at its input terminals?
2.38 The input impedance of a 31-cm long lossless transmission line of unknown characteristic impedance was measured
at 1 MHz. With the line terminated in a short circuit, the
measurement yielded an input impedance equivalent to an
inductor with inductance of 0.064 µ H, and when the line was
open-circuited, the measurement yielded an input impedance
equivalent to a capacitor with capacitance of 40 pF. Find Z0 of
the line, the phase velocity, and the relative permittivity of the
insulating material.
∗ 2.39 A 75 Ω resistive load is preceded by a λ /4 section
of a 50 Ω lossless line, which itself is preceded by another
λ /4 section of a 100 Ω line. What is the input impedance?
Compare your result with that obtained through two successive
applications of Module 2.5.
2.40 A 100 MHz FM broadcast station uses a 300 Ω transmission line between the transmitter and a tower-mounted
half-wave dipole antenna. The antenna impedance is 73 Ω. You
are asked to design a quarter-wave transformer to match the
antenna to the line.
(a) Determine the electrical length and characteristic impedance of the quarter-wave section.
(b) If the quarter-wave section is a two-wire line with
D = 2.5 cm and the wires are embedded in polystyrene
with εr = 2.6, determine the physical length of the
quarter-wave section and the radius of the two wire
conductors.
eg = 300 V and Zg = 50 Ω is con2.42 A generator with V
nected to a load ZL = 75 Ω through a 50 Ω lossless line of
length l = 0.15λ .
∗ (a) Compute Z , the input impedance of the line at the
in
generator end.
(b) Compute Iei and Vei .
(c) Compute the time-average power delivered to the line,
Pin = 12 Re[Vei Iei∗ ].
eL , IeL , and the time-average power delivered to
(d) Compute V
the load, PL = 12 Re[VeL IeL∗ ]. How does Pin compare with PL ?
Explain.
(e) Compute the time-average power delivered by the generator, Pg , and the time-average power dissipated in Zg . Is
conservation of power satisfied?
2.43 If the two-antenna configuration shown in Fig. P2.43 is
connected to a generator with Veg = 250 V and Zg = 50 Ω, how
much average power is delivered to each antenna?
∗ 2.44 For the circuit shown in Fig. P2.44, calculate the
average incident power, the average reflected power, and the
average power transmitted into the infinite 100 Ω line. The
λ /2 line is lossless and the infinitely long line is slightly lossy.
(Hint: The input impedance of an infinitely long line is equal
to its characteristic impedance so long as α 6= 0.)
2.45
The circuit shown in Fig. P2.45 consists of a
100 Ω lossless transmission line terminated in a load with
ZL = (50 + j100) Ω. If the peak value of the load voltage was
measured to be |VeL | = 12 V, determine:
∗ (a) the time-average power dissipated in the load,
(b) the time-average power incident on the line, and
(c) the time-average power reflected by the load.
126
CHAPTER 2
λ/2
50 Ω
λ/2
A
+
Zin
250 V
−
Lin
C
TRANSMISSION LINES
ZL1 = 75 Ω
(Antenna 1)
e2
Line 1
B
D
Lin
Generator
e3
λ/2
ZL2 = 75 Ω
(Antenna 2)
Figure P2.43 Antenna configuration for Problem 2.43.
2.46 An antenna with a load impedance
50 Ω
ZL = (75 + j25) Ω
λ/2
Z0 = 50 Ω
2V
Z1 = 100 Ω
8
+
−
i
Pav
t
Pav
r
Pav
Figure P2.44 Circuit for Problem 2.44.
is connected to a transmitter through a 50 Ω lossless transmission line. If under matched conditions (50 Ω load) the
transmitter can deliver 20 W to the load, how much power can
it deliver to the antenna? Assume that Zg = Z0 .
Section 2-10: Smith Chart
2.47 Use the Smith chart to find the reflection coefficient
corresponding to a load impedance of
(a) ZL = 3Z0
∗ (b) Z = (2 − j2)Z
L
0
(c) ZL = − j2Z0
(d) ZL = 0 (short circuit)
Rg
~
Vg
2.48 Repeat Problem 2.47 using Module 2.6.
2.49 Use the Smith chart to find the normalized load impedance corresponding to a reflection coefficient of
+
Z0 = 100 Ω
ZL = (50 + j100) Ω
−
(a) Γ = 0.5
(b) Γ = 0.5 60◦
Figure P2.45 Circuit for Problem 2.45.
(c) Γ = −1
(d) Γ = 0.3 −30◦
(e) Γ = 0
(f) Γ = j
PROBLEMS
127
l1 = 3λ/8
C
B
Zin
Z01 = 100 Ω
l2 = 5λ/8
A
Z02 = 50 Ω
ZL = (75 − j50) Ω
Figure P2.50 Circuit for Problem 2.50.
∗ 2.50 Use the Smith chart to determine the input impedance
Zin of the two-line configuration shown in Fig. P2.50.
2.51 Repeat Problem 2.50 using Module 2.6.
∗ 2.52 On a lossless transmission line terminated in a load
ZL = 100 Ω, the standing-wave ratio was measured to be 2.5.
Use the Smith chart to find the two possible values of Z0 .
2.53 A lossless 50 Ω transmission line is terminated in a
load with ZL = (50 + j25) Ω. Use the Smith chart to find the
following:
(a) The reflection coefficient Γ.
∗ (b) The standing-wave ratio.
(c) The input impedance at 0.35λ from the load.
(d) The input admittance at 0.35λ from the load.
(e) The shortest line length for which the input impedance is
purely resistive.
(f) The position of the first voltage maximum from the load.
2.59 A 75 Ω lossless line is 0.6λ long. If S = 1.8 and
θr = −60◦, use the Smith chart to find |Γ|, ZL , and Zin .
2.60 Repeat Problem 2.59 using Module 2.6.
∗ 2.61 Using a slotted line on a 50 Ω air-spaced lossless
line, the following measurements were obtained: S = 1.6 and
|Ve |max occurred only at 10 cm and 24 cm from the load. Use
the Smith chart to find ZL .
2.62 At an operating frequency of 5 GHz, a 50 Ω lossless
coaxial line with insulating material having a relative permittivity εr = 2.25 is terminated in an antenna with an impedance
ZL = 150 Ω. Use the Smith chart to find Zin . The line length is
30 cm.
Section 2-11: Impedance Matching
∗ 2.63 A 50 Ω lossless line 0.6λ long is terminated in a load
with ZL = (50 + j25) Ω. At 0.3λ from the load, a resistor with
resistance R = 30 Ω is connected as shown in Fig. P2.63. Use
the Smith chart to find Zin .
2.54 Repeat Problem 2.53 using Module 2.6.
∗ 2.55 A lossless 50 Ω transmission line is terminated in a
short circuit. Use the Smith chart to determine:
(a) The input impedance at a distance 2.3λ from the load.
Zin
Z0 = 50 Ω
30 Ω
Z0 = 50 Ω
ZL
(b) The distance from the load at which the input admittance
is Yin = − j0.04 S.
2.56 Repeat Problem 2.55 using Module 2.6.
∗ 2.57 Use the Smith chart to find y if z = 1.5 − j0.7.
L
L
2.58 A lossless 100 Ω transmission line 3λ /8 in length is
terminated in an unknown impedance. If the input impedance
is Zin = − j2.5 Ω,
(a) Use the Smith chart to find ZL .
(b) Verify your results using Module 2.6.
0.3λ
0.3λ
ZL = (50 + j25) Ω
Figure P2.63 Circuit for Problem 2.63.
2.64 Use Module 2.7 to design a quarter-wavelength transformer to match a load with ZL = (100 − j200) Ω to a 50 Ω
line.
128
CHAPTER 2
2.65 Use Module 2.7 to design a quarter-wavelength transformer to match a load with ZL = (50 + j10) Ω to a 100 Ω
line.
2.66 A 200 Ω transmission line is to be matched to a
computer terminal with ZL = (50 − j25) Ω by inserting an
appropriate reactance in parallel with the line. If f = 800 MHz
and εr = 4, determine the location nearest to the load at which
inserting:
(a) A capacitor can achieve the required matching, and the
value of the capacitor.
(b) An inductor can achieve the required matching, and the
value of the inductor.
TRANSMISSION LINES
2.74 A 25 Ω antenna is connected to a 75 Ω lossless
transmission line. Reflections back toward the generator can
be eliminated by placing a shunt impedance Z at a distance l
from the load (Fig. P2.74). Determine the values of Z and l.
l=?
B
Z0 = 75 Ω
A
Z=?
ZL = 25 Ω
2.67 Repeat Problem 2.66 using Module 2.8.
2.68 A 50 Ω lossless line is to be matched to an antenna with
ZL = (75 − j20) Ω using a shorted stub. Use the Smith chart
to determine the stub length and distance between the antenna
and stub.
Figure P2.74 Circuit for Problem 2.74.
∗ 2.69 Repeat Problem 2.68 for a load with
ZL = (100 + j50) Ω.
Section 2-12: Transients on Transmission Lines
2.70 Repeat Problem 2.68 using Module 2.9.
2.71 Repeat Problem 2.69 using Module 2.9.
2.72 Determine Zin of the feed line shown in Fig. P2.72. All
lines are lossless with Z0 = 50 Ω.
Z1 = (50 + j50) Ω
λ
0.3
0.3λ
2.75 Generate a bounce diagram for the voltage V (z,t) for
a 1-m long lossless line characterized by Z0 = 50 Ω and
up = 2c/3 (where c is the velocity of light) if the line is fed
by a step voltage applied at t = 0 by a generator circuit with
Vg = 60 V and Rg = 100 Ω. The line is terminated in a load
RL = 25 Ω. Use the bounce diagram to plot V (t) at a point
midway along the length of the line from t = 0 to t = 25 ns.
2.76 Repeat Problem 2.75 for the current I(z,t) on the line.
Z1
2.77 In response to a step voltage, the voltage waveform
shown in Fig. P2.77 was observed at the sending end of a
lossless transmission line with Rg = 50 Ω, Z0 = 50 Ω, and
εr = 2.25. Determine:
Zin
0.7
(a) the generator voltage,
λ
Z2
(b) the length of the line, and
(c) the load impedance.
Z2 = (50 − j50) Ω
Figure P2.72 Network for Problem 2.72.
∗ 2.73 Repeat Problem 2.72 for the case where all three transmission lines are λ /4 in length.
∗ 2.78 In response to a step voltage, the voltage waveform
shown in Fig. P2.78 was observed at the sending end of a
shorted line with Z0 = 50 Ω and εr = 4. Determine Vg , Rg ,
and the line length.
PROBLEMS
129
2.81 For the circuit of Problem 2.80, generate a bounce
diagram for the current and plot its time history at the middle
of the line.
V(0, t)
∗ 2.82 In response to a step voltage, the voltage waveform
5V
shown in Fig. P2.82 was observed at the midpoint of a lossless
transmission line with Z0 = 50 Ω and up = 2 × 108 m/s.
Determine: (a) the length of the line, (b) ZL , (c) Rg , and (d) Vg .
3V
t
0
6 μs
V(l/2, t)
Figure P2.77 Voltage waveform for Problems 2.77 and 2.79.
12 V
V(0, t)
12 V
0
15
3
21
t (μs)
9
−3 V
3V
0
7 μs
0.75 V
Figure P2.82 Circuit for Problem 2.82.
t
14 μs
Figure P2.78 Voltage waveform of Problem 2.78.
2.79 Suppose the voltage waveform shown in Fig. P2.77
was observed at the sending end of a 50 Ω transmission line
in response to a step voltage introduced by a generator with
Vg = 15 V and an unknown series resistance Rg . The line is
1 km in length, its velocity of propagation is 1 × 108 m/s, and
it is terminated in a load RL = 100 Ω.
(a) Determine Rg .
(b) Explain why the drop in level of V (0,t) at t = 6 µ s cannot
be due to reflection from the load.
(c) Determine the shunt resistance Rf and location of the fault
responsible for the observed waveform.
2.80 A generator circuit with Vg = 200 V and Rg = 25 Ω
was used to excite a 75 Ω lossless line with a rectangular
pulse of duration τ = 0.4 µ s. The line is 200 m long, its up =
2 × 108 m/s, and it is terminated in a load RL = 125 Ω.
(a) Synthesize the voltage pulse exciting the line as the sum
of two step functions, Vg1 (t) and Vg2 (t).
(b) For each voltage step function, generate a bounce diagram
for the voltage on the line.
(c) Use the bounce diagrams to plot the total voltage at the
sending end of the line.
(d) Confirm the result of part (c) by applying Module 2.10.
2.83 The transmission-line circuit of Fig. P2.83 is excited by
a 1-ns long pulse that starts at t = 0. The line is 1 m long and
the phase velocity of the line is 2 × 108 m/s. Generate:
(a) a plot of the voltage across the line at t = 6 ns,
(b) a plot of the voltage across the line at t = 12 ns,
(c) a plot of the voltage as a function of time at the sending
end of the line, and
(d) a plot of the voltage as a function of time at the midpoint
of the line.
+
5V
0
1 ns
_
10 W
Z0 = 50 Ω
Figure P2.83 Circuit for Problem 2.83.
100 W
Chapter
3
Vector Analysis
Chapter Contents
3-1
3-2
3-3
3-4
TB5
3-5
3-6
TB6
3-7
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Overview, 131
Basic Laws of Vector Algebra, 131
Orthogonal Coordinate Systems, 137
Transformations between Coordinate Systems, 143
Gradient of a Scalar Field, 147
Global Positioning System, 151
Divergence of a Vector Field, 153
Curl of a Vector Field, 157
X-Ray Computed Tomography, 159
Laplacian Operator, 162
Chapter 3 Summary, 163
Problems, 164
1. Use vector algebra in Cartesian, cylindrical, and spherical
coordinate systems.
2. Transform vectors between the three primary coordinate
systems.
3. Calculate the gradient of a scalar function and the divergence and curl of a vector function in any of the three
primary coordinate systems.
4. Apply the divergence theorem and Stokes’s theorem.
130
Overview
A
In our examination of wave propagation on a transmission
A = aˆ A
line in Chapter 2, the primary quantities we worked with were
voltage, current, impedance, and power. Each of these is a
scalar quantity, meaning that it can be completely specified by
its magnitude if it is a positive real number or by its magnitude
and phase angle if it is a negative or a complex number (a
negative number has a positive magnitude and a phase angle
of π (rad)). This chapter is concerned with vectors. A vector
has a magnitude and a direction. The speed of an object is a
scalar, whereas its velocity is a vector.
Starting in the next chapter and throughout the succeeding
chapters in this book, the primary electromagnetic quantities we deal with are the electric and magnetic fields, E
and H. These, and many other related quantities, are vectors. Vector analysis provides the mathematical tools necessary for expressing and manipulating vector quantities in
an efficient and convenient manner. To specify a vector in
three-dimensional space, it is necessary to specify its components along each of the three directions.
aˆ
1
Figure 3-1 Vector A = âA has magnitude A = |A| and points
in the direction of unit vector â = A/A.
z
3
2
1 zˆ
1
yˆ
1
2
3
y
xˆ
2
3
x
(a) Base vectors
◮ Several types of coordinate systems are used in the
study of vector quantities, the most common being the
Cartesian (or rectangular), cylindrical, and spherical systems. A particular coordinate system is usually chosen to
best suit the geometry of the problem under consideration. ◭
z
Az
A
Az
Vector algebra governs the laws of addition, subtraction,
and “multiplication” of vectors. The rules of vector algebra and
vector representation in each of the aforementioned orthogonal
coordinate systems (including vector transformation between
them) are two of the three major topics treated in this chapter.
The third topic is vector calculus, which encompasses the
laws of differentiation and integration of vectors, the use of
special vector operators (gradient, divergence, and curl), and
the application of certain theorems that are particularly useful
in the study of electromagnetics, most notably the divergence
and Stokes’s theorems.
Ay
Ax
x
(b) Components of A
Figure 3-2 Cartesian coordinate system: (a) base vectors x̂, ŷ,
and ẑ and (b) components of vector A.
The unit vector â has a magnitude of one (|â| = 1) and points
from A’s tail or anchor to its head or tip. From Eq. (3.1),
3-1 Basic Laws of Vector Algebra
A vector is a mathematical object that resembles an arrow.
Vector A in Fig. 3-1 has magnitude (or length) A = |A| and
unit vector â:
A = â|A| = âA.
y
Ar
â =
A
A
= .
|A|
A
(3.2)
In the Cartesian (or rectangular) coordinate system shown
in Fig. 3-2(a), the x, y, and z coordinate axes extend along
(3.1)
131
132
CHAPTER 3 VECTOR ANALYSIS
directions of the three mutually perpendicular unit vectors x̂,
ŷ, and ẑ, which are also called base vectors. The vector A in
Fig. 3-2(b) may be decomposed as
A = x̂Ax + ŷAy + ẑAz ,
Since A is a nonnegative scalar, only the positive root applies.
From Eq. (3.2), the unit vector â is
(3.5)
Occasionally, we use the shorthand notation A = (Ax , Ay , Az ) to
denote a vector with components Ax , Ay , and Az in a Cartesian
coordinate system.
3-1.1 Equality of Two Vectors
Two vectors A and B are equal if they have equal magnitudes
and identical unit vectors. Thus, if
A = âA = x̂Ax + ŷAy + ẑAz ,
(3.6a)
B = b̂B = x̂Bx + ŷBy + ẑBz ,
(3.6b)
then A = B if and only if A = B and â = b̂, which requires that
Ax = Bx , Ay = By , and Az = Bz .
◮ Equality of two vectors does not necessarily imply that
they are identical; in Cartesian coordinates, two displaced
parallel vectors of equal magnitude and pointing in the
same direction are equal, but they are identical only if they
lie on top of one another. ◭
B
(a) Parallelogram rule
The sum of two vectors A and B is a vector
C = x̂Cx + ŷCy + ẑCz ,
A
B
(b) Head-to-tail rule
Figure 3-3 Vector addition by (a) the parallelogram rule and
(b) the head-to-tail rule.
given by
C = A + B = (x̂Ax + ŷAy + ẑAz ) + (x̂Bx + ŷBy + ẑBz )
= x̂(Ax + Bx ) + ŷ(Ay + By ) + ẑ(Az + Bz)
= x̂Cx + ŷCy + ẑCz ,
(3.7)
with Cx = Ax + Bx , etc.
◮ Vector addition is commutative:
C = A + B = B + A.
(3.8)
Graphically, vector addition can be accomplished by either the
parallelogram or the head-to-tail rule (Fig. 3-3). Vector C is
the diagonal of the parallelogram with sides A and B. With the
head-to-tail rule, we may either add A to B or B to A. When
A is added to B, it is repositioned so that its tail starts at the
tip of B while keeping its length and direction unchanged. The
sum vector C starts at the tail of B and ends at the tip of A.
Subtraction of vector B from vector A is equivalent to the
addition of A to negative B. Thus,
D = A − B = A + (−B)
= x̂(Ax − Bx ) + ŷ(Ay − By ) + ẑ(Az − Bz ).
(3.9)
Graphically, the same rules used for vector addition are also
applicable to vector subtraction; the only difference is that the
arrowhead of (−B) is drawn on the opposite end of the line
segment representing the vector B (i.e., the tail and head are
interchanged).
3-1.3
3-1.2 Vector Addition and Subtraction
C
C
(3.3)
where Ax , Ay , and Az are A’s scalar components along the
x-, y-, and z axes, respectively. The component Az is equal to
the perpendicular projection of A onto the z axis, and similar
definitions apply to Ax and Ay . Application of the Pythagorean
theorem—first to the right triangle in the x–y plane to express
the hypotenuse Ar in terms of Ax and Ay and then again to the
vertical right triangle with sides Ar and Az and hypotenuse A—
yields the following expression for the magnitude of A:
q
A = |A| = + A2x + A2y + A2z .
(3.4)
A x̂Ax + ŷAy + ẑAz
â = = q
.
A
+
A2x + A2y + A2z
A
Position and Distance Vectors
The position vector of a point P in space is the vector from the
origin to P. Assuming points P1 and P2 are at (x1 , y1 , z1 ) and
(x2 , y2 , z2 ) in Fig. 3-4, their position vectors are
−→
R1 = OP1 = x̂x1 + ŷy1 + ẑz1 ,
(3.10a)
3-1 BASIC LAWS OF VECTOR ALGEBRA
133
Scalar or Dot Product
z
z2
The scalar (or dot) product of two co-anchored vectors A
and B, denoted A · B and pronounced “A dot B,” is defined
geometrically as the product of the magnitude of A and the
scalar component of B along A, or vice versa. Thus,
P1 = (x1, y1, z1)
z1
R12
R1
P2 = (x2, y2, z2)
A · B = AB cos θAB ,
R2
y2
y1
O
y
x1
x2
x
−−→
Figure 3-4 Distance vector R12 = P1 P2 = R2 − R1 , where R1
(3.14)
where θAB is the angle between A and B (Fig. 3-5) measured
from the tail of A to the tail of B. Angle θAB is assumed to be
in the range 0 ≤ θAB ≤ 180◦. The scalar product of A and B
yields a scalar whose magnitude is less than or equal to the
products of their magnitudes (equality holds when θAB = 0)
and whose sign is positive if 0 < θAB < 90◦ and negative if
90◦ < θAB < 180◦. When θAB = 90◦ , A and B are orthogonal,
and their dot product is zero. The quantity A cos θAB is the
scalar component of A along B. Similarly, B cos θBA is the
scalar component of B along A.
and R2 are the position vectors of points P1 and P2 , respectively.
−→
R2 = OP2 = x̂x2 + ŷy2 + ẑz2 ,
where point O is the origin.
The distance vector from P1 to P2 is defined as
−−→
R12 = P1 P2 = R2 − R1
= x̂(x2 − x1 ) + ŷ(y2 − y1 ) + ẑ(z2 − z1 ),
(3.10b)
The dot product obeys both the commutative and
distributive properties of multiplication:
A · B = B · A,
(commutative property)
A ·(B + C) = A · B + A · C.
(3.11)
and the distance d between P1 and P2 equals the magnitude
of R12 :
d = |R12 | = [(x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 ]1/2 . (3.12)
Note that the first and second subscripts of R12 denote the
locations of its tail and head, respectively (Fig. 3-4).
(3.15b)
(distributive property)
The commutative property follows from Eq. (3.14) and the fact
that θAB = θBA . The distributive property expresses the fact that
the scalar component of the sum of two vectors along a third
one equals the sum of their respective scalar components.
The dot product of a vector with itself gives
A · A = |A|2 = A2 ,
3-1.4 Vector Multiplication
There exist three types of products in vector calculus: the
simple product, the scalar (or dot) product, and the vector (or
cross) product.
B
θAB
The multiplication of a vector by a scalar is called a simple
product. The product of the vector A = âA by a scalar k results
in a vector B with magnitude B = kA and direction the same
as A. That is, b̂ = â. In Cartesian coordinates,
B = kA = âkA = x̂(kAx ) + ŷ(kAy ) + ẑ(kAz )
(3.13)
(3.16)
A
θAB
θBA
Simple Product
= x̂ Bx + ŷ By + ẑBz .
(3.15a)
B
(a)
A
θBA
(b)
Figure 3-5 The angle θAB is the angle between A and B,
measured from A to B between vector tails. The dot product
is positive if 0 ≤ θAB < 90◦ , as in (a), and it is negative if
90◦ < θAB ≤ 180◦ , as in (b).
134
CHAPTER 3 VECTOR ANALYSIS
which implies that
z
√
+
A = |A| = A · A .
A × B = nˆ AB sin θAB
(3.17)
Also, θAB can be determined from
A·B
√
.
θAB = cos−1 √
+
A·A + B·B
B
nˆ
θAB
(3.18)
y
Since the base vectors x̂, ŷ, and ẑ are each orthogonal to the
other two, it follows that
A
x
(a) Cross product
x̂ · x̂ = ŷ · ŷ = ẑ · ẑ = 1,
(3.19a)
x̂ · ŷ = ŷ · ẑ = ẑ · x̂ = 0.
A×B
(3.19b)
If A = (Ax , Ay , Az ) and B = (Bx , By , Bz ), then
B
A · B = (x̂Ax + ŷAy + ẑAz ) ·(x̂Bx + ŷBy + ẑBz ).
(3.20)
Use of Eqs. (3.19a) and (3.19b) in Eq. (3.20) leads to
A · B = Ax Bx + Ay By + Az Bz .
A
Figure 3-6 Cross product A × B points in the direction n̂,
Vector or Cross Product
The vector (or cross) product of two vectors A and B, denoted
× B and pronounced “A cross B,” yields a vector defined as
A×
× B = n̂ AB sin θAB ,
A×
(3.22)
where n̂ is a unit vector normal to the plane containing A
and B (Fig. 3-6(a)). The magnitude of the cross product,
AB| sin θAB |, equals the area of the parallelogram defined by the
two vectors. The direction of n̂ is governed by the right-hand
rule (Fig. 3-6(b)): n̂ points in the direction of the right thumb
when the fingers rotate from A to B through the angle θAB .
Note that, since n̂ is perpendicular to the plane containing A
× B is perpendicular to both vectors A and B.
and B, A×
The cross product is anticommutative and distributive:
× B = −B×
×A
A×
(anticommutative).
(b) Right-hand rule
(3.21)
(3.23a)
which is perpendicular to the plane containing A and B and
defined by the right-hand rule.
by A and (B + C) equals the sum of those formed by (A and B)
and (A and C):
× (B + C) = A×
× B + A×
× C,
A×
(3.23b)
(distributive)
The cross product of a vector with itself vanishes. That is,
× A = 0.
A×
(3.24)
From the definition of the cross product given by Eq. (3.22),
it is easy to verify that the base vectors x̂, ŷ, and ẑ of
the Cartesian coordinate system obey the right-hand cyclic
relations:
× ŷ = ẑ,
x̂×
× ẑ = x̂,
ŷ×
× x̂ = ŷ.
ẑ×
(3.25)
Note the cyclic order (xyzxyz . . .). Also,
The anticommutative property follows from the application of
the right-hand rule to determine n̂. The distributive property
follows from the fact that the area of the parallelogram formed
× ẑ = 0.
x̂ × x̂ = ŷ × ŷ = ẑ×
(3.26)
3-1 BASIC LAWS OF VECTOR ALGEBRA
135
p
√
22 + 32 + 32 = 22 ,
√
A
â = = (x̂2 + ŷ3 + ẑ3)/ 22 .
A
If A = (Ax , Ay , Az ) and B = (Bx , By , Bz ), then use of Eqs. (3.25)
and (3.26) leads to
× B = (x̂Ax + ŷAy + ẑAz )×
× (x̂Bx + ŷBy + ẑBz )
A×
= x̂(Ay Bz − AzBy ) + ŷ(Az Bx − Ax Bz )
+ ẑ(Ax By − Ay Bx ).
(b) The angle β between A and the y axis is obtained from
(3.27)
The cyclical form of the result given by Eq. (3.27) allows us to
express the cross product in the form of a determinant:
×B =
A×
x̂
Ax
Bx
ŷ
Ay
By
A = |A| =
ẑ
Az .
Bz
A · ŷ = |A||ŷ| cos β = A cos β ,
or
−1
β = cos
(c)
(3.28)
A · ŷ
A
−1
= cos
3
√
22
= 50.2◦.
B = x̂(1 − 2) + ŷ(−2 − 3) + ẑ(2 − 3) = −x̂ − ŷ5 − ẑ.
(d)
Example 3-1:
−1
θAB = cos
Vectors and Angles
In Cartesian coordinates, vector A points from the origin to
point P1 = (2, 3, 3), and vector B is directed from P1 to point
P2 = (1, −2, 2). Find:
(a)
(b)
(c)
(d)
(e)
vector A, its magnitude A, and unit vector â,
the angle between A and the y axis,
vector B,
the angle θAB between A and B, and
perpendicular distance from the origin to vector B.
Solution: (a) Vector A is given by the position vector of
P1 = (2, 3, 3) (Fig. 3-7). Thus,
A = x̂2 + ŷ3 + ẑ3,
θAB
2
O
Example 3-2: Cross Product
β
1
2
x
Figure 3-7 Geometry of Example 3-1.
x̂
2
0
ŷ
ẑ
−1 3
2 −3
= x̂((−1) × (−3) − 3 × 2) − ŷ(2 × (−3) − 3 × 0)
A
1
–2
−→
| OP3 | = |A| sin(180◦ − θAB )
√
= 22 sin(180◦ − 145.1◦) = 2.68.
×B =
A×
P1 = (2, 3, 3)
P3
(e) The perpendicular distance between the origin and vector B
−→
is the distance | OP3 | shown in Fig. 3-7. From right triangle
OP1 P3 ,
Solution: (a) Application of Eq. (3.28) gives
3
B
A·B
−1 (−2 − 15 − 3)
√ √
= cos
= 145.1◦.
|A||B|
22 27
Given vectors A = x̂2 − ŷ + ẑ3 and B = ŷ2 − ẑ3, compute
× B, (b) ŷ × B, and (c) (ŷ×
× B) · A.
(a) A×
z
P2 = (1, –2, 2)
+ ẑ(2 × 2 − (−1 × 0))
3
y
= −x̂3 + ŷ6 + ẑ4.
× B = ŷ×
× (ŷ2 − ẑ3) = −x̂3.
(b) ŷ×
(c) (ŷ × B) · A = −x̂3 ·(x̂2 − ŷ + ẑ3) = −6.
Find the distance vector between
P1 = (1, 2, 3) and P2 = (−1, −2, 3) in Cartesian coordinates.
−−→
Answer: P1 P2 = −x̂2 − ŷ4. (See EM .)
Exercise 3-1:
136
CHAPTER 3 VECTOR ANALYSIS
Exercise 3-2: Find the angle θAB between vectors A and B
result is a scalar. A scalar triple product obeys the cyclic order:
of Example 3-1 from the cross product between them.
Answer: θAB = 145.1◦. (See
EM
× A) = C ·(A×
× B).
A ·(B × C) = B ·(C×
.)
Exercise 3-3: Find the angle between vector B of Exam-
ple 3-1 and the z axis.
Answer: 101.1◦. (See
EM
.)
The equalities hold as long as the cyclic order (ABCABC . . .) is
preserved. The scalar triple product of vectors A = (Ax , Ay , Az ),
B = (Bx , By , Bz ), and C = (Cx ,Cy ,Cz ) can be expressed in the
form of a 3 × 3 determinant:
Exercise 3-4: Vectors A and B lie in the y–z plane and
× C) =
A ·(B×
both have the same magnitude of 2 (Fig. E3.4). Determine
× B.
(a) A · B and (b) A×
z
Ax
Bx
Cx
Ay
By
Cy
Az
Bz
Cz
.
Vector Triple Product
2 30◦
A
2
The vector triple product involves the cross product of a vector
with the cross product of two others, such as
y
× (B × C).
A×
x
× B = x̂ 3.46. (See
Answer: (a) A · B = −2; (b) A×
(3.31)
Since each cross product yields a vector, the result of a vector
triple product is also a vector. The vector triple product does
not obey the associative law. That is,
Figure E3.4
EM
.)
Exercise 3-5: If A · B = A · C, does it follow that B = C?
EM
(3.30)
The validity of Eqs. (3.29) and (3.30) can be verified by
expanding A, B, and C in component form and carrying out
the multiplications.
B
Answer: No. (See
(3.29)
.)
3-1.5 Scalar and Vector Triple Products
When three vectors are multiplied, not all combinations of dot
and cross products are meaningful. For example, the product
× (B · C)
A×
does not make sense because B · C is a scalar, and the cross
product of the vector A with a scalar is not defined under the
rules of vector algebra. Other than the product of the form
A(B · C), the only two meaningful products of three vectors
are the scalar triple product and the vector triple product.
× (B × C) 6= (A×
× B)×
× C,
A×
which means that it is important to specify which cross multiplication is to be performed first. By expanding the vectors A,
B, and C in component form, it can be shown that
× (B×
× C) = B(A · C) − C(A · B),
A×
(3.33)
which is known as the “bac-cab” rule.
Example 3-3: Vector Triple Product
Given A = x̂ − ŷ + ẑ2, B = ŷ + ẑ, and C = −x̂2 + ẑ3, find
× B)×
× C and compare it with A×
× (B × C).
(A×
Solution:
Scalar Triple Product
The dot product of a vector with the cross product of two other
vectors is called a scalar triple product, so named because the
(3.32)
×B =
A×
x̂
1
0
ŷ ẑ
−1 2
1 1
= −x̂3 − ŷ + ẑ
3-2 ORTHOGONAL COORDINATE SYSTEMS
and
× B)×
×C =
(A×
x̂
ŷ
−3 −1
−2 0
ẑ
1
3
= −x̂3 + ŷ7 − ẑ2.
A similar procedure gives A × (B × C) = x̂2 + ŷ4 + ẑ. The
fact that the results of two vector triple products are different
demonstrates the inequality stated in Eq. (3.32).
When are two vectors equal
and when are they identical?
137
Why do we need more than one coordinate system? Whereas
a point in space has the same location and an object has the
same shape regardless of which coordinate system is used
to describe them, the solution of a practical problem can be
greatly facilitated by the choice of a coordinate system that
best fits the geometry under consideration. The following subsections examine the properties of each of the aforementioned
orthogonal systems, and Section 3-3 describes how a point or
vector may be transformed from one system to another.
Concept Question 3-1:
When is the position vector of a
point identical to the distance vector between two points?
Concept Question 3-2:
Concept Question 3-3:
If A · B = 0, what is θAB ?
Concept Question 3-4:
× B = 0, what is θAB ?
If A×
Concept Question 3-5:
Is A(B · C) a vector triple prod-
uct?
Concept Question 3-6:
that B = C?
If A · B = A · C, does it follow
3-2.1 Cartesian Coordinates
The Cartesian coordinate system was introduced in Section 3-1
to illustrate the laws of vector algebra. Instead of repeating
these laws for the Cartesian system, we summarize them in
Table 3-1. Differential calculus involves the use of differential lengths, areas, and volumes. In Cartesian coordinates, a
differential length vector (Fig. 3-8) is expressed as
dl = x̂ dlx + ŷ dly + ẑ dlz = x̂ dx + ŷ dy + ẑ dz,
(3.34)
where dlx = dx is a differential length along x̂, and similar
interpretations apply to dly = dy and dlz = dz.
A differential area vector ds is a vector with magnitude ds
equal to the product of two differential lengths (such as dly
and dlz ) and direction specified by a unit vector along the third
3-2 Orthogonal Coordinate Systems
A three-dimensional coordinate system allows us to uniquely
specify locations of points in space and the magnitudes and
directions of vectors. Coordinate systems may be orthogonal
or nonorthogonal.
dsz = zˆ dx dy
z
dy
dx
dsy = yˆ dx dz
◮ An orthogonal coordinate system is one in which
coordinates are measured along locally mutually perpendicular axes. ◭
dz
dz
dv = dx dy dz
dl
dsx = xˆ dy dz
Nonorthogonal systems are very specialized and seldom used
in solving practical problems. Many orthogonal coordinate
systems have been devised, but the most commonly used are
• the Cartesian (also called rectangular),
• the cylindrical, and
• the spherical coordinate system.
dy
y
dx
x
Figure 3-8 Differential length, area, and volume in Cartesian
coordinates.
138
CHAPTER 3 VECTOR ANALYSIS
Table 3-1 Summary of vector relations.
Cartesian
Coordinates
Cylindrical
Coordinates
Spherical
Coordinates
x, y, z
r, φ , z
R, θ , φ
x̂Ax + ŷAy + ẑAz
q
+
A2x + A2y + A2z
r̂Ar + φ̂φAφ + ẑAz
q
+
A2r + Aφ2 + A2z
R̂AR + θ̂θAθ + φ̂φAφ
q
+
A2R + Aθ2 + Aφ2
x̂x1 + ŷy1 + ẑz1 ,
for P(x1 , y1 , z1 )
r̂r1 + ẑz1 ,
for P(r1 , φ1 , z1 )
R̂R1 ,
for P(R1 , θ1 , φ1 )
x̂ · x̂ = ŷ · ŷ = ẑ · ẑ = 1
x̂ · ŷ = ŷ · ẑ = ẑ · x̂ = 0
x̂ × ŷ = ẑ
ŷ × ẑ = x̂
ẑ × x̂ = ŷ
φ = ẑ · ẑ = 1
r̂ · r̂ = φ̂φ · φ̂
r̂ · φ̂φ = φ̂φ · ẑ = ẑ · r̂ = 0
r̂ × φ̂φ = ẑ
φ × ẑ = r̂
φ̂
φ
ẑ × r̂ = φ̂
θ · θ̂
θ = φ̂φ · φ̂φ = 1
R̂ · R̂ = θ̂
θ · φ̂
φ = φ̂φ · R̂ = 0
R̂ · θ̂θ = θ̂
× θ̂
θ = φ̂φ
R̂×
θ × φ̂
φ = R̂
θ̂
φ × R̂ = θ̂θ
φ̂
Coordinate variables
Vector representation A =
Magnitude of A
Position vector
|A| =
−→
OP =
1
Base vector properties
Dot product
Cross product
Differential length
A·B =
Ax Bx + Ay By + Az Bz
x̂
Ax
Bx
×B =
A×
dl =
ŷ
Ay
By
ẑ
Az
Bz
r̂
Ar
Br
φ̂φ
Aφ
Bφ
ẑ
Az
Bz
AR BR + Aθ Bθ + Aφ Bφ
R̂
AR
BR
θ̂θ
Aθ
Bθ
φ̂φ
Aφ
Bφ
x̂ dx + ŷ dy + ẑ dz
r̂ dr + φ̂φr d φ + ẑ dz
R̂ dR + θ̂θR d θ + φ̂φR sin θ d φ
dsx = x̂ dy dz
dsy = ŷ dx dz
dsz = ẑ dx dy
dsr = r̂r d φ dz
φ dr dz
dsφ = φ̂
dsz = ẑr dr d φ
dsR = R̂R2 sin θ d θ d φ
dsθ = θ̂θR sin θ dR d φ
dsφ = φ̂φR dR d θ
dx dy dz
r dr d φ dz
R2 sin θ dR d θ d φ
Differential surface areas
Differential volume d υ =
direction (such as x̂). Thus, for a differential area vector in the
y–z plane,
dsx = x̂ dly dlz = x̂ dy dz
Ar Br + Aφ Bφ + Az Bz
(y–z plane),
(3.35a)
with the subscript on ds denoting its direction. Similarly,
dsy = ŷ dx dz
(x–z plane),
(3.35b)
dsz = ẑ dx dy
(x–y plane).
(3.35c)
A differential volume equals the product of all three differential lengths:
d υ = dx dy dz.
(3.36)
3-2.2 Cylindrical Coordinates
The cylindrical coordinate system is useful for solving
problems involving structures with cylindrical symmetry, such
as calculating the capacitance per unit length of a coaxial
transmission line. In the cylindrical coordinate system, the
location of a point in space is defined by three variables: r,
φ , and z (Fig. 3-9). The coordinate r is the radial distance in
the x–y the azimuth angle measured from the positive x axis,
and z is as previously defined in the Cartesian coordinate
system. Their ranges are 0 ≤ r < ∞, 0 ≤ φ < 2π , and
−∞ < z < ∞. Point P(r1 , φ1 , z1 ) in Fig. 3-9 is located at the
intersection of three surfaces. These are the cylindrical surface
defined by r = r1 , the vertical half-plane defined by φ = φ1
(which extends outwardly from the z axis), and the horizontal
plane defined by z = z1 .
φ , and ẑ
◮ The mutually perpendicular base vectors are r̂, φ̂
with r̂ pointing away from the origin along r, φ̂φ pointing
in a direction tangential to the cylindrical surface, and ẑ
pointing along the vertical. Unlike the Cartesian system,
where base vectors x̂, ŷ, and ẑ are independent of the
location of P, both r̂ and φ̂φ are functions of φ in the
cylindrical system. ◭
3-2 ORTHOGONAL COORDINATE SYSTEMS
139
z
z = z1 plane
P = (r1, φ1, z1)
z1
R1
r = r1 cylinder
O
φ1
y
r1
zˆ
ˆ
φ
φ = φ1 plane
rˆ
x
Figure 3-9 Point P(r1 , φ1 , z1 ) in cylindrical coordinates; r1 is the radial distance from the origin in the x–y plane, φ1 is the angle in the
x–y plane measured from the x axis toward the y axis, and z1 is the vertical distance from the x–y plane.
The base unit vectors obey the following right-hand cyclic
relations:
× φ̂φ = ẑ,
r̂×
φ̂φ × ẑ = r̂,
× r̂ = φ̂
φ,
ẑ×
(3.37)
and like all unit vectors, r̂ · r̂ = φ̂φ · φ̂φ = ẑ · ẑ = 1, and
× r̂ = φ̂φ × φ̂φ = ẑ×
× ẑ = 0.
r̂×
In cylindrical coordinates, a vector is expressed as
A = â|A| = r̂Ar + φ̂φAφ + ẑAz ,
(3.38)
where Ar , Aφ , and Az are the components of A along the r̂, φ̂φ,
and ẑ directions. The magnitude of A is obtained by applying
Eq. (3.17), which gives
q
√
+
|A| = A · A = + A2r + A2φ + A2z .
(3.39)
−→
The position vector OP shown in Fig. 3-9 has components
along r and z only. Thus,
−→
R1 = OP = r̂r1 + ẑz1 .
(3.40)
The dependence of R1 on φ1 is implicit through the dependence of r̂ on φ1 . Hence, when using Eq. (3.40) to denote the
position vector of point P(r1 , φ1 , z1 ), it is necessary to specify
that r̂ is at φ1 .
Figure 3-10 shows a differential volume element in cylindrical coordinates. The differential lengths along r̂, φ̂φ, and ẑ
are
dlr = dr,
dlφ = r d φ ,
dlz = dz.
(3.41)
Note that the differential length along φ̂φ is r d φ , not just d φ .
The differential length dl in cylindrical coordinates is given by
dl = r̂ dlr + φ̂φ dlφ + ẑ dlz = r̂ dr + φ̂φr d φ + ẑ dz.
(3.42)
As was stated previously for the Cartesian coordinate system,
the product of any pair of differential lengths is equal to the
magnitude of a vector differential surface area with a surface
normal pointing along the direction of the third coordinate.
Thus,
dsr = r̂ dlφ dlz = r̂r d φ dz
(φ –z cylindrical surface),
(3.43a)
φ dr dz
dsφ = φ̂φ dlr dlz = φ̂
(r–z plane),
(3.43b)
dsz = ẑ dlr dlφ = ẑr dr d φ
(r–φ plane).
(3.43c)
140
CHAPTER 3 VECTOR ANALYSIS
z
z
P1 = (0, 0, h)
dsz = zˆ r dr dφ
dz
r dφ
dr
aˆ
dsφ = ϕˆ dr dz
h
dv = r dr dφ dz
dz
dsr = rˆ r dφ dz
A
O
φ0 r0
O
P2 = (r0, φ0, 0)
y
φ
x
r
x
r dφ
dr
Figure 3-10
y
Figure 3-11 Geometry of Example 3-4.
Differential areas and volume in cylindrical
coordinate system, which is not true. The ambiguity can be
resolved by specifying that A passes through a point whose
φ = φ0 .
coordinates.
The differential volume is the product of the three differential
lengths,
d υ = dlr dlφ dlz = r dr d φ dz.
(3.44)
These properties of the cylindrical coordinate system are
summarized in Table 3-1.
Example 3-4:
Distance Vector in
Cylindrical Coordinates
Example 3-5: Cylindrical Area
Find the area of a cylindrical surface described by r = 5,
30◦ ≤ φ ≤ 60◦ , and 0 ≤ z ≤ 3 (Fig. 3-12).
z
Find an expression for the unit vector of vector A shown in
Fig. 3-11 in cylindrical coordinates.
z=3
r=5
Solution: In triangle OP1 P2 ,
−→
−→
OP2 = OP1 +A.
Hence,
−→ −→
A = OP2 − OP1 = r̂r0 − ẑh,
and
â =
A
r̂r0 − ẑh
=q
.
|A|
r02 + h2
We note that the expression for A is independent of φ0 . This
implies that all vectors from point P1 to any point on the circle
defined by r = r0 in the x–y plane are equal in the cylindrical
y
60°
30°
x
Figure 3-12 Cylindrical surface of Example 3-5.
3-2 ORTHOGONAL COORDINATE SYSTEMS
141
Module 3.1 Points and Vectors Examine the relationships between Cartesian coordinates (x, y) and cylindrical
coordinates (r, φ ) for points and vectors.
Solution: The prescribed surface is shown in Fig. 3-12. Use
of Eq. (3.43a) for a surface element with constant r gives
S=r
Z 60◦
φ =30◦
dφ
Z 3
dz = 5φ
z=0
π /3
π /6
3
z =
0
5π
.
2
Note that φ had to be converted to radians before evaluating
the integration limits.
Exercise 3-6: A circular cylinder of radius r = 5 cm is
concentric with the z axis and extends between z = −3 cm
and z = 3 cm. Use Eq. (3.44) to find the cylinder’s volume.
Answer: 471.2 cm3 . (See
EM
.)
distance from the origin to the point, describes a sphere of
radius R centered at the origin. The zenith angle θ is measured
from the positive z axis and it describes a conical surface with
its apex at the origin, and the azimuth angle φ is the same
as in cylindrical coordinates. The ranges of R, θ , and φ are
0 ≤ R < ∞, 0 ≤ θ ≤ π , and 0 ≤ φ < 2π . The base vectors R̂,
θ̂θ, and φ̂φ obey the right-hand cyclic relations:
θ = φ̂φ,
R̂ × θ̂
φ = R̂,
θ̂θ × φ̂
φ̂φ × R̂ = θ̂θ.
(3.45)
A vector with components AR , Aθ , and Aφ is written as
A = â|A| = R̂AR + θ̂θAθ + φ̂φAφ ,
(3.46)
and its magnitude is
3-2.3 Spherical Coordinates
In the spherical coordinate system, the location of a point
in space is uniquely specified by the variables R, θ , and φ
(Fig. 3-13). The range coordinate R, which measures the
|A| =
q
√
+
A · A = + A2R + A2θ + A2φ .
The position vector of point P(R1 , θ1 , φ1 ) is simply
−→
R1 = OP = R̂R1 ,
(3.47)
(3.48)
142
CHAPTER 3 VECTOR ANALYSIS
As shown in Fig. 3-14, the differential lengths along R̂, θ̂θ,
and φ̂φ are
z
dlR = dR,
φ̂
R1
θˆ
dlφ = R sin θ d φ .
(3.49)
Hence, the expressions for the vector differential length dl, the
vector differential surface ds, and the differential volume d υ
are
ˆ
R
θ = θ1
conical
surface
dlθ = R d θ ,
P = (R1, θ1, φ1)
dl = R̂ dlR + θ̂θ dlθ + φ̂φ dlφ
θ1
y
= R̂ dR + θ̂θR d θ + φ̂φR sin θ d φ ,
(3.50a)
dsR = R̂ dlθ dlφ = R̂R2 sin θ d θ d φ
(3.50b)
φ1
(θ –φ spherical surface),
dsθ = θ̂θ dlR dlφ = θ̂θR sin θ dR d φ
ˆ
φ
(3.50c)
(R–φ conical surface),
x
dsφ = φ̂φ dlR dlθ = φ̂φR dR d θ
Figure 3-13 Point P(R1 , θ1 , φ1 ) in spherical coordinates.
(R–θ plane),
(3.50d)
d υ = dlR dlθ dlφ = R2 sin θ dR d θ d φ .
(3.50e)
These relations are summarized in Table 3-1.
z
Example 3-6:
R sin θ dφ
Surface Area in Spherical
Coordinates
dν = R2 sin θ dR dθ dφ
The spherical strip shown in Fig. 3-15 is a section of a sphere
of radius 3 cm. Find the area of the strip.
dR
R
θ
R dθ
z
dθ
y
φ
30 o
60 o
dφ
3c
m
x
y
Figure 3-14 Differential volume in spherical coordinates.
x
while keeping in mind that R̂ is implicitly dependent on θ1
and φ1 .
Figure 3-15 Spherical strip of Example 3-6.
3-3 TRANSFORMATIONS BETWEEN COORDINATE SYSTEMS
Solution: Use of Eq. (3.50b) for the area of an elemental
spherical area with constant radius R gives
S = R2
Z 60◦
θ =30◦
sin θ d θ
= 9(− cos θ )
Z 2π
dφ
φ =0
60◦
30◦
φ
2π
section, we shall establish the relations between the variables
(x, y, z) of the Cartesian system, (r, φ , z) of the cylindrical system, and (R, θ , φ ) of the spherical system. These relations will
then be used to transform expressions for vectors expressed in
any one of the three systems into expressions applicable in the
other two.
(cm2 )
0
= 18π (cos30◦ − cos60◦) = 20.7 cm2 .
Example 3-7:
143
Charge in a Sphere
A sphere of radius 2 cm contains a volume charge density ρv
given by
ρv = 4 cos2 θ
(C/m3 ).
3-3.1 Cartesian to Cylindrical Transformations
Point P in Fig. 3-16 has Cartesian coordinates (x, y, z) and
cylindrical coordinates (r, φ , z). Both systems share the coordinate z, and the relations between the other two pairs of
coordinates can be obtained from the geometry in Fig. 3-16.
They are
y
p
r = + x2 + y2 ,
,
(3.51)
φ = tan−1
x
and the inverse relations are
Find the total charge Q contained in the sphere.
Solution:
Q=
=
Z
υ
ρv d υ
φ =0 θ =0 R=0
3
0
32
=
× 10−6
3
sin θ cos2 θ d θ d φ
cos3 θ
−
3
2π
64
dφ
× 10−6
9
0
128π
× 10−6 = 44.68
=
9
=
(3.52)
r̂ · ŷ = sin φ ,
(3.53a)
φ̂φ · ŷ = cos φ .
φ · x̂ = − sin φ ,
φ̂
(3.53b)
To express r̂ in terms of x̂ and ŷ, we write r̂ as
0
Z 2π 0
r̂ · x̂ = cos φ ,
(4 cos2 θ )R2 sin θ dR d θ d φ
Z 2π Z π 3 2×10−2
R
0
y = r sin φ .
Next, with the help of Fig. 3-17, which shows the directions of
the unit vectors x̂, ŷ, r̂, and φ̂φ in the x–y plane, we obtain the
relations:
Z 2π Z π Z 2×10−2
=4
x = r cos φ ,
r̂ = x̂a + ŷb,
π
dφ
0
Z
(3.54)
where a and b are unknown transformation coefficients. The
dot product r̂ · x̂ gives
r̂ · x̂ = x̂ · x̂a + ŷ · x̂b = a.
(3.55)
(µ C).
z
Note that the limits on R were converted to meters prior to
evaluating the integral on R.
φ r
123
y = r sin φ
1
The position of a given point in space of course does not
depend on the choice of coordinate system. That is, its location
is the same irrespective of which specific coordinate system is
used to represent it. The same is true for vectors. Nevertheless,
certain coordinate systems may be more useful than others in
solving a given problem, so it is essential that we have the tools
to “translate” the problem from one system to another. In this
2
3-3 Transformations between Coordinate
Systems
3
P(x, y, z)
z
y
x = r cos φ
x
Figure 3-16 Interrelationships between Cartesian coordinates
(x, y, z) and cylindrical coordinates (r, φ , z).
144
CHAPTER 3 VECTOR ANALYSIS
and, conversely,
y
ϕ
Ax = Ar cos φ − Aφ sin φ ,
ϕ̂
r
Ay = Ar sin φ + Aφ cos φ .
(3.59a)
(3.59b)
yˆ
−ϕˆ
The transformation relations given in this and the following
two subsections are summarized in Table 3-2.
ϕ rˆ
xˆ
x
Example 3-8:
Figure 3-17 Interrelationships between base vectors (x̂, ŷ) and
Cartesian to Cylindrical
Transformations
(r̂, φ̂φ).
Comparison of Eq. (3.55) with Eq. (3.53a) yields a = cos φ .
Similarly, application of the dot product r̂ · ŷ to Eq. (3.54) gives
b = sin φ . Hence,
r̂ = x̂ cos φ + ŷsin φ .
(3.56a)
Solution: For point P1 , x = 3, y = −4, and z = 3. Using
Eq. (3.51), we have
p
y
r = + x2 + y2 = 5, φ = tan−1 = −53.1◦ = 306.9◦,
x
and z remains unchanged. Hence, P1 = P1 (5, 306.9◦, 3) in
cylindrical coordinates.
The cylindrical components of vector A = r̂Ar + φ̂φAφ + ẑAz
can be determined by applying Eqs. (3.58a) and (3.58b):
Repetition of the procedure for φ̂φ leads to
φ = −x̂ sin φ + ŷ cos φ .
φ̂
Given point P1 = (3, −4, 3) and vector A = x̂2 − ŷ3 + ẑ4
defined in Cartesian coordinates, express P1 and A in cylindrical coordinates and evaluate A at P1 .
(3.56b)
The third base vector ẑ is the same in both coordinate systems.
By solving Eqs. (3.56a) and (3.56b) simultaneously for x̂
and ŷ, we obtain the following inverse relations:
x̂ = r̂ cos φ − φ̂φ sin φ ,
(3.57a)
ŷ = r̂ sin φ + φ̂φ cos φ .
(3.57b)
The relations given by Eqs. (3.56a) to (3.57b) are not only
φ ), and
useful for transforming the base vectors (x̂, ŷ) into (r̂, φ̂
vice versa, they also can be used to transform the components
of a vector expressed in either coordinate system into its
corresponding components expressed in the other system. For
example, a vector A = x̂Ax + ŷAy + ẑAz in Cartesian coordinates can be described by A = r̂Ar + φ̂φAφ + ẑAz in cylindrical
coordinates by applying Eqs. (3.56a) and (3.56b). That is,
Ar = Ax cos φ + Ay sin φ ,
(3.58a)
Aφ = −Ax sin φ + Ay cos φ ,
(3.58b)
Ar = Ax cos φ + Ay sin φ = 2 cos φ − 3 sin φ ,
Aφ = −Ax sin φ + Ay cos φ = −2 sin φ − 3 cos φ ,
Az = 4.
Hence,
A = r̂(2 cos φ − 3 sin φ ) − φ̂φ(2 sin φ + 3 cos φ ) + ẑ4.
At point P, φ = 306.9◦, which gives
A = r̂ 3.60 − φ̂φ 0.20 + ẑ 4.
3-3.2
Cartesian to Spherical Transformations
From Fig. 3-18, we obtain the following relations between
the Cartesian coordinates (x, y, z) and the spherical coordinates
(R, θ , φ ):
p
(3.60a)
R = + x2 + y2 + z2 ,
#
"p
+ 2
x + y2
,
(3.60b)
θ = tan−1
z
y
φ = tan−1
.
(3.60c)
x
3-3 TRANSFORMATIONS BETWEEN COORDINATE SYSTEMS
145
Table 3-2 Coordinate transformation relations.
Transformation
Cartesian to
cylindrical
Coordinate Variables
p
r = + x2 + y2
φ = tan−1 (y/x)
z=z
x = r cos φ
y = r sin φ
z=z
p
R = + x2 + y2 + z2
Cylindrical to
Cartesian
Cartesian to
spherical
θ = tan−1 [
p
+
x2 + y2 /z]
φ = tan−1 (y/x)
x = R sin θ cos φ
Spherical to
Cartesian
y = R sin θ sin φ
z = R cos θ
√
+
R = r2 + z2
θ = tan−1 (r/z)
φ =φ
Cylindrical to
spherical
r = R sin θ
φ =φ
z = R cos θ
Spherical to
cylindrical
Unit Vectors
Vector Components
r̂ = x̂ cos φ + ŷ sin φ
φ = −x̂ sin φ + ŷ cos φ
φ̂
ẑ = ẑ
Ar = Ax cos φ + Ay sin φ
Aφ = −Ax sin φ + Ay cos φ
Az = Az
x̂ = r̂ cos φ − φ̂φ sin φ
ŷ = r̂ sin φ + φ̂φ cos φ
ẑ = ẑ
Ax = Ar cos φ − Aφ sin φ
Ay = Ar sin φ + Aφ cos φ
Az = Az
R̂ = x̂ sin θ cos φ
+ ŷ sin θ sin φ + ẑ cos θ
θ = x̂ cos θ cos φ
θ̂
+ ŷ cos θ sin φ − ẑ sin θ
φ = −x̂ sin φ + ŷ cos φ
φ̂
AR = Ax sin θ cos φ
+ Ay sin θ sin φ + Az cos θ
Aθ = Ax cos θ cos φ
+ Ay cos θ sin φ − Az sin θ
Aφ = −Ax sin φ + Ay cos φ
x̂ = R̂ sin θ cos φ
+ θ̂θ cos θ cos φ − φ̂φ sin φ
ŷ = R̂ sin θ sin φ
+ θ̂θ cos θ sin φ + φ̂φ cos φ
ẑ = R̂ cos θ − θ̂θ sin θ
Ax = AR sin θ cos φ
+ Aθ cos θ cos φ − Aφ sin φ
Ay = AR sin θ sin φ
+ Aθ cos θ sin φ + Aφ cos φ
Az = AR cos θ − Aθ sin θ
R̂ = r̂ sin θ + ẑ cos θ
θ = r̂ cos θ − ẑ sin θ
θ̂
φ = φ̂φ
φ̂
AR = Ar sin θ + Az cos θ
Aθ = Ar cos θ − Az sin θ
Aφ = Aφ
r̂ = R̂ sin θ + θ̂θ cos θ
φ = φ̂φ
φ̂
ẑ = R̂ cos θ − θ̂θ sin θ
Ar = AR sin θ + Aθ cos θ
Aφ = Aφ
Az = AR cos θ − Aθ sin θ
The converse relations are
z
θ
zˆ
ˆ
R
(π/2 – θ)
R
θ
x = R sin θ cos φ ,
(3.61a)
y = R sin θ sin φ ,
(3.61b)
z = R cos θ .
(3.61c)
The unit vector R̂ lies in the r̂–ẑ plane. Hence, it can be
expressed as a linear combination of r̂ and ẑ as
r̂
z = R cos θ
R̂ = r̂a + ẑb,
(3.62)
y
φ
x = r cos φ
r
φ̂
y = r sin φ
x
rˆ
Figure 3-18 Interrelationships between (x, y, z) and (R, θ , φ ).
where a and b are transformation coefficients. Since r̂ and ẑ
are mutually perpendicular,
R̂ · r̂ = a,
R̂ · ẑ = b.
(3.63a)
(3.63b)
From Fig. 3-18, the angle between R̂ and r̂ is the complement
of θ and that between R̂ and ẑ is θ . Hence, a = R̂ · r̂ = sin θ
146
CHAPTER 3 VECTOR ANALYSIS
and b = R̂ · ẑ = cos θ . Upon inserting these expressions for a
and b in Eq. (3.62) and replacing r̂ with Eq. (3.56a), we have
R̂ = x̂ sin θ cos φ + ŷ sin θ sin φ + ẑ cos θ .
Aθ = (x + y) cos θ cos φ + (y − x) cos θ sin φ − z sin θ ,
(3.64a)
A similar procedure can be followed to obtain the expression
for θ̂θ:
θ = x̂ cos θ cos φ + ŷcos θ sin φ − ẑ sin θ .
θ̂
Similarly,
(3.64b)
Aφ = −(x + y) sin φ + (y − x) cos φ ,
and following the procedure used with AR , we obtain
Aθ = 0,
Hence,
A = R̂AR + θ̂θAθ + φ̂φAφ = R̂R − φ̂φR sin θ .
Finally φ̂φ is given by Eq. (3.56b) as
φ = −x̂ sin φ + ŷ cos φ .
φ̂
(3.64c)
Equations (3.64a) through (3.64c) can be solved simultaneθ, φ̂
φ):
ously to give the expressions for (x̂, ŷ, ẑ) in terms of (R̂, θ̂
x̂ = R̂ sin θ cos φ + θ̂θ cos θ cos φ − φ̂φ sin φ ,
(3.65a)
ŷ = R̂ sin θ sin φ + θ̂θ cos θ sin φ + φ̂φ cos φ ,
(3.65b)
ẑ = R̂ cos θ − θ̂θ sin θ .
(3.65c)
Equations (3.64a) to (3.65c) also can be used to transform
Cartesian components (Ax , Ay , Az ) of vector A into their spherical counterparts (AR , Aθ , Aφ ), and vice versa, by replacing
θ, φ̂
φ) with (Ax , Ay , Az , AR , Aθ , Aφ ).
(x̂, ŷ, ẑ, R̂, θ̂
Example 3-9:
Cartesian to Spherical
Transformation
Express vector A = x̂(x + y) + ŷ(y − x) + ẑz in spherical
coordinates.
Solution: Using the transformation relation for AR given in
Table 3-2, we have
AR = Ax sin θ cos φ + Ay sin θ sin φ + Az cos θ
= (x + y) sin θ cos φ + (y − x) sin θ sin φ + z cos θ .
Using the expressions for x, y, and z given by Eq. (3.61c), we
have
AR = (R sin θ cos φ + R sin θ sin φ ) sin θ cos φ
+ (R sin θ sin φ −R sin θ cos φ ) sin θ sin φ + R cos2 θ
= R sin2 θ (cos2 φ + sin2 φ ) + R cos2 θ
= R sin2 θ + R cos2 θ = R.
Aφ = −R sin θ .
3-3.3
Cylindrical to Spherical Transformations
Transformations between cylindrical and spherical coordinates
can be realized by combining the transformation relations
of the preceding two subsections. The results are given in
Table 3-2.
3-3.4
Distance between Two Points
In Cartesian coordinates, the distance d between two points
P1 = (x1 , y1 , z1 ) and P2 = (x2 , y2 , z2 ) is given by Eq. (3.12) as
d = |R12 | = [(x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 ]1/2 .
(3.66)
Upon using Eq. (3.52) to convert the Cartesian coordinates of
P1 and P2 into their cylindrical equivalents, we have
d = (r2 cos φ2 − r1 cos φ1 )2
+ (r2 sin φ2 − r1 sin φ1 )2 + (z2 − z1 )2
1/2
1/2
.
= r22 +r12 − 2r1 r2 cos(φ2 −φ1 )+(z2 −z1 )2
(3.67)
(cylindrical)
A similar transformation using Eqs. (3.61a) through (3.61c)
leads to an expression for d in terms of the spherical coordinates of P1 and P2 :
d = R22 + R21 − 2R1R2 [cos θ2 cos θ1
+ sin θ1 sin θ2 cos(φ2 − φ1 )]
(spherical)
1/2
.
(3.68)
3-4 GRADIENT OF A SCALAR FIELD
Example 3-10:
Vector Component
At a given point in space, vectors A and B are given in
cylindrical coordinates by
A = r̂2 + φ̂φ3 − ẑ,
147
Why is it that the base vectors
(x̂, ŷ, ẑ) are independent of the location of a point, but r̂
and φ̂φ are not?
Concept Question 3-8:
What are the cyclic relations
for the base vectors in (a) Cartesian coordinates, (b) cylindrical coordinates, and (c) spherical coordinates?
Concept Question 3-9:
B = r̂ + ẑ.
Determine (a) the scalar component of B, or projection, in the
direction of A, (b) the vector component of B in the direction
of A, and (c) the vector component of B perpendicular to A.
How is the position vector of
a point in cylindrical coordinates related to its position
vector in spherical coordinates?
Concept Question 3-10:
√
Exercise 3-7: Point P = (2 3, π /3, −2) is given in cylin-
drical coordinates. Express P in spherical coordinates.
B
Answer: P = (4, 2π /3, π /3). (See
EM
.)
D
C
A
Figure 3-19 Vectors A, B, C, and D of Example 3-10.
Exercise 3-8: Transform vector
A = x̂(x + y) + ŷ(y − x) + ẑz
from Cartesian to cylindrical coordinates.
φr + ẑz. (See
Answer: A = r̂r − φ̂
EM
.)
Solution: (a) Let us denote the scalar component of B in the
direction of A as C, as shown in Fig. 3-19. Thus,
C = B · â = B ·
(r̂2 + φ̂φ3 − ẑ) 2 − 1
A
= (r̂ + ẑ) · √
= √ = 0.267.
|A|
4+9+1
14
(b) The vector component of B in the direction of A is given
by the product of the scalar component C and the unit vector â:
C = âC =
(r̂2 + φ̂φ3 − ẑ)
A
√
× 0.267
C=
|A|
14
= r̂0.143 + φ̂φ0.214 − ẑ0.071.
(c) The vector component of B perpendicular to A is equal
to B minus C:
D = B − C = (r̂ + ẑ) − (r̂0.143 + φ̂φ0.214 − ẑ0.071)
= r̂0.857 − φ̂φ0.214 + ẑ0.929.
Concept Question 3-7:
coordinate system?
Why do we use more than one
3-4 Gradient of a Scalar Field
When dealing with a scalar physical quantity whose magnitude
depends on a single variable, such as the temperature T as
a function of height z, the rate of change of T with height
can be described by the derivative dT/dz. However, if T is
also a function of x and y, its spatial rate of change becomes
more difficult to describe because we now have to deal with
three separate variables. The differential change in T along x,
y, and z can be described in terms of the partial derivatives
of T with respect to the three coordinate variables, but it is
not immediately obvious as to how we should combine the
three partial derivatives so as to describe the spatial rate of
change of T along a specified direction. Furthermore, many
of the quantities we deal with in electromagnetics are vectors;
therefore, both their magnitudes and directions may vary with
spatial position. To this end, we introduce three fundamental
operators to describe the differential spatial variations of
scalars and vectors: the gradient, divergence, and curl operators. The gradient operator applies to scalar fields and is the
subject of the present section. The other two operators, which
apply to vector fields, are discussed in succeeding sections.
148
CHAPTER 3 VECTOR ANALYSIS
Equation (3.71) then can be expressed as
z
P2 = (x + dx, y + dy, z + dz)
dT = ∇T · dl.
dy
dx
dl
(3.73)
The symbol ∇ is called the del or gradient operator and is
defined as
dz
P1 = (x, y, z)
∇ = x̂
y
∂
∂
∂
+ ŷ + ẑ
∂x
∂y
∂z
(Cartesian).
(3.74)
x
Figure 3-20 Differential distance vector dl between points P1
and P2 .
Suppose that T1 = T (x, y, z) is the temperature at point
P1 = (x, y, z) in some region of space, and
◮ Whereas the gradient operator itself has no physical
meaning, it attains a physical meaning once it operates on
a scalar quantity, and the result of the operation is a vector
with magnitude equal to the maximum rate of change of
the physical quantity per unit distance and pointing in the
direction of maximum increase. ◭
T2 = T (x + dx, y + dy, z + dz)
With dl = âl dl, where âl is the unit vector of dl, the directional
derivative of T along âl is
is the temperature at a nearby point P2 = (x+dx, y+dy, z+dz)
(Fig. 3-20). The differential distances dx, dy, and dz are the
components of the differential distance vector dl. That is,
dT
= ∇T · âl .
dl
dl = x̂ dx + ŷ dy + ẑ dz.
(3.69)
From differential calculus, the temperature difference between
points P1 and P2 , dT = T2 − T1 , is
dT =
∂T
∂T
∂T
dx +
dy +
dz.
∂x
∂y
∂z
(3.70)
Because dx = x̂ · dl, dy = ŷ · dl, and dz = ẑ · dl, Eq. (3.70) can
be rewritten as
∂T
∂T
∂T
· dl + ŷ · dl + ẑ · dl
∂x
∂y
∂z
∂T
∂T
∂T
· dl.
+ ŷ
+ ẑ
= x̂
∂x
∂y
∂z
(3.75)
We can find the difference (T2 − T1 ), where T1 = T (x1 , y1 , z1 )
and T2 = T (x2 , y2 , z2 ) are the values of T at points
P1 = (x1 , y1 , z1 )
and
P2 = (x2 , y2 , z2 )
not necessarily infinitesimally close to one another, by integrating both sides of Eq. (3.73). Thus,
T2 − T1 =
Z P2
P1
∇T · dl.
(3.76)
dT = x̂
Example 3-11: Directional Derivative
(3.71)
The vector inside the square brackets in Eq. (3.71) relates the
change in temperature dT to a vector change in direction dl.
This vector is called the gradient of T (or grad T for short)
and denoted ∇T :
∇T = grad T = x̂
∂T
∂T
∂T
+ ŷ
+ ẑ
.
∂x
∂y
∂z
(3.72)
Find the directional derivative of T = x2 + y2 z along direction
x̂2 + ŷ3 − ẑ2 and evaluate it at (1, −1, 2).
Solution: First, we find the gradient of T :
∂
∂
∂
(x2 + y2 z) = x̂2x + ŷ2yz + ẑy2 .
∇T = x̂ + ŷ + ẑ
∂x
∂y
∂z
We denote l as the given direction,
l = x̂2 + ŷ3 − ẑ2.
3-4 GRADIENT OF A SCALAR FIELD
149
Hence, the gradient operator in cylindrical coordinates can be
expressed as
Its unit vector is
âl =
l
x̂2 + ŷ3 − ẑ2
x̂2 + ŷ3 − ẑ2
√
.
=√
=
2
2
2
|l|
17
2 +3 +2
∇ = r̂
Application of Eq. (3.75) gives
x̂2 + ŷ3 − ẑ2
dT
2
√
= ∇T · âl = (x̂2x + ŷ2yz + ẑy ) ·
dl
17
=
4x + 6yz − 2y2
√
.
17
At (1, −1, 2),
∂
∂
1 ∂
+ ẑ ,
+ φ̂φ
∂r
r ∂φ
∂z
(3.82)
A similar procedure leads to the expression for the gradient in
spherical coordinates:
∇ = R̂
dT
dl
(cylindrical)
4 − 12 − 2 −10
=√ .
= √
17
17
(1,−1,2)
∂
1
1 ∂
∂
+ φ̂φ
.
+ θ̂θ
∂R
R ∂θ
R sin θ ∂ φ
(3.83)
(spherical)
3-4.1 Gradient Operator in Cylindrical and
Spherical Coordinates
3-4.2 Properties of the Gradient Operator
Even though Eq. (3.73) was derived using Cartesian coordinates, it should have counterparts in other coordinate systems.
To convert Eq. (3.72) into cylindrical coordinates (r, φ , z), we
start by restating the coordinate relations
p
y
tan φ = .
(3.77)
r = x2 + y2 ,
x
For any two scalar functions U and V , the following relations
apply:
(1) ∇(U + V ) = ∇U + ∇V ,
(3.84a)
(2) ∇(UV ) = U ∇V + V ∇U,
(3.84b)
(3) ∇V n = nV n−1 ∇V ,
(3.84c)
for any n.
From differential calculus,
∂T
∂T ∂r ∂T ∂φ ∂T ∂z
=
+
+
.
∂x
∂r ∂x ∂φ ∂x
∂z ∂x
Example 3-12: Calculating the Gradient
(3.78)
Since z is orthogonal to x and ∂ z/∂ x = 0, the last term in
Eq. (3.78) vanishes. From the coordinate relations given by
Eq. (3.77), it follows that
x
∂r
= cos φ ,
=p
2
∂x
x + y2
∂φ
1
= − sin φ .
∂x
r
(3.79a)
(3.79b)
Hence,
∂ T sin φ ∂ T
∂T
.
= cos φ
−
∂x
∂r
r ∂φ
∂T
∂T
1 ∂T
+ φ̂φ
.
+ ẑ
∂r
r ∂φ
∂z
(a) V1 = 24V0 cos (π y/3)sin (2π z/3) at (3, 2, 1) in Cartesian
coordinates,
(b) V2 = V0 e−2r sin 3φ at (1, π /2, 3) in cylindrical coordinates,
(c) V3 = V0 (a/R)cos 2θ at (2a, 0, π ) in spherical coordinates.
Solution: (a) Using Eq. (3.72) for ∇,
∂ V1
∂ V1
∂ V1
+ ŷ
+ ẑ
∂x
∂y
∂z
2π z
2π z
πy
πy
sin
+ ẑ16π V0 cos
cos
= −ŷ8π V0 sin
3
3
3
3
πy
πy
2π z
2π z
.
= 8π V0 −ŷ sin
sin
+ ẑ2 cos
cos
3
3
3
3
∇V1 = x̂
(3.80)
This expression can be used to replace the coefficient of x̂ in
Eq. (3.72), and a similar procedure can be followed to obtain
an expression for ∂ T /∂ y in terms of r and φ . If, in addition, we
use the relations x̂ = r̂ cos φ − φ̂φ sin φ and ŷ = r̂ sin φ + φ̂φ cos φ
[from Eqs. (3.57a) and (3.57b)], then Eq. (3.72) becomes
∇T = r̂
Find the gradient of each of the following scalar functions and
then evaluate it at the given point.
(3.81)
At (3, 2, 1),
2π
2π
∇V1 = 8π V0 −ŷ sin2
= π V0 [−ŷ6 + ẑ4] .
+ ẑ2 cos2
3
3
150
CHAPTER 3 VECTOR ANALYSIS
Module 3.2 Gradient Select a scalar function f (x, y, z), evaluate its gradient, and display both in an appropriate 2-D plane.
(b) The function V2 is expressed in terms of cylindrical
variables. Hence, we need to use Eq. (3.82) for ∇:
1 ∂
∂
∂
V0 e−2r sin 3φ
∇V2 = r̂ + φ̂φ
+ ẑ
∂r
r ∂φ
∂z
= −r̂2V0 e−2r sin 3φ + φ̂φ(3V0 e−2r cos 3φ )/r
3 cos 3φ
V0 e−2r .
= −r̂2 sin 3φ + φ̂φ
r
At (1, π /2, 3), r = 1 and φ = π /2. Hence,
3π
3π
∇V2 = −r̂2 sin
V0 e−2 = r̂2V0 e−2 = r̂0.27V0.
+ φ̂φ3 cos
2
2
(c) As V3 is expressed in spherical coordinates, we apply
Eq. (3.83) to V3 :
a
1
1 ∂
∂
∂
cos 2θ
V0
+ φ̂φ
+ θ̂θ
∇V3 = R̂
∂R
R ∂θ
R sin θ ∂ φ
R
V0 a
2V0 a
cos 2θ − θ̂θ 2 sin 2θ
R2
R
V0 a
= −[R̂ cos 2θ + θ̂θ2 sin 2θ ] 2 .
R
= −R̂
At (2a, 0, π ), R = 2a and θ = 0, which yields
∇V3 = −R̂
V0
.
4a
Exercise 3-9: Given V = x2 y + xy2 + xz2 , (a) find the
gradient of V , and (b) evaluate it at (1, −1, 2).
(a) ∇V = x̂(2xy + y2 + z2 ) + ŷ(x2 + 2xy)
+ ẑ2xz, (b) ∇V (1,−1,2) = x̂3 − ŷ + ẑ4. (See EM .)
Answer:
Exercise 3-10: Find the directional derivative of
V = rz2 cos 2φ along the direction of A = r̂2 − ẑ and
evaluate it at (1, π /2, 2).
√
Answer: (dV/dl) (1,π /2,2) = −4/ 5 . (See EM .)
3-4 GRADIENT OF A SCALAR FIELD
151
Technology Brief 5:
Global Positioning System
The Global Positioning System (GPS), initially developed in the 1980s by the U.S. Department of Defense
as a navigation tool for military use, has evolved into
a system with numerous civilian applications, including
vehicle tracking, aircraft navigation, map displays in automobiles and hand-held cell phones (Fig. TF5-1), and
topographic mapping. The overall GPS comprises three
segments. The space segment consists of 24 satellites
(Fig. TF5-2), each circling Earth every 12 hours at an
orbital altitude of about 12,000 miles and transmitting
continuous coded time signals. All satellite transmitters
broadcast coded messages at two specific frequencies:
1.57542 GHz and 1.22760 GHz. The user segment
consists of hand-held or vehicle-mounted receivers that
determine their own locations by receiving and processing multiple satellite signals. The third segment is a
network of five ground stations distributed around the
world that monitor the satellites and provide them with
updates on their precise orbital information.
◮ GPS provides a location inaccuracy of about
30 m both horizontally and vertically, but it can be
improved to within 1 m by differential GPS. (See
final section.) ◭
Figure TF5-1 iPhone map feature.
Principle of Operation
The triangulation technique allows the determination
of the location (x0 , y0 , z0 ) of any object in 3-D space
from knowledge of the distances d1 , d2 , and d3 between
that object and three other independent points in space
of known locations (x1 , y1 , z1 ) to (x3 , y3 , z3 ). In GPS, the
distances are established by measuring the times it
takes the signals to travel from the satellites to the GPS
receivers, and then multiplying them by the speed of
light c = 3 × 108 m/s. Time synchronization is achieved
by using atomic clocks. The satellites use very precise
clocks, accurate to 3 nanoseconds (3 × 10−9 s), but
receivers use less accurate, inexpensive, ordinary quartz
clocks. Consequently, the receiver clock may have an
unknown time offset error t0 relative to the satellite
clocks. To correct for the time error of a GPS receiver,
a signal from a fourth satellite is needed.
Figure TF5-2 GPS nominal satellite constellation. Four
satellites in each plane with 20,200 km altitudes and a 55◦
inclination.
152
CHAPTER 3 VECTOR ANALYSIS
SAT 3
(x3, y3, z3)
SAT 4
(x4, y4, z4)
d3
SAT 2
(x2, y2, z2)
d2
SAT 1
(x1, y1, z1)
d4
d1
Time delay
(x0, y0, z0)
Receiver Code
Satellite Code
Figure TF5-3
Automobile GPS receiver at location
(x0 , y0 , z0 ).
The GPS receiver of the automobile in Fig. TF5-3 is at
distances d1 to d4 from the GPS satellites. Each satellite
sends a message identifying its orbital coordinates
(x1 , y1 , z1 ) for satellite 1, and so on for the other satellites,
together with a binary-coded sequence common to all
satellites. The GPS receiver generates the same binary
sequence (Fig. TF5-3), and by comparing its code with
the one received from satellite 1, it determines the time
t1 corresponding to travel time over the distance d1 . A
similar process applies to satellites 2 to 4, leading to four
equations:
d12 = (x1 − x0 )2 + (y1 − y0 )2 + (z1 − z0 )2 = c2 [(t1 + t0 )]2 ,
d22 = (x2 − x0 )2 + (y2 − y0 )2 + (z2 − z0 )2 = c2 [(t2 + t0 )]2 ,
d32 = (x3 − x0 )2 + (y3 − y0 )2 + (z3 − z0 )2 = c2 [(t3 + t0 )]2 ,
d42 = (x4 − x0 )2 + (y4 − y0 )2 + (z4 − z0 )2 = c2 [(t4 + t0 )]2 .
The four satellites report their coordinates (x1 , y1 , z1 ) to
(x4 , y4 , z4 ) to the GPS receiver, and the time delays t1 to
t4 are measured directly by the receiver. The unknowns
are (x0 , y0 , z0 ), the coordinates of the GPS receiver, and
the time offset of its clock t0 . Simultaneous solution of the
four equations provides the desired location information.
Differential GPS
The 30 m GPS position inaccuracy is attributed to several
factors, including time-delay errors (due to the difference between the speed of light and the actual signal
speed in the troposphere) that depend on the receiver’s
location on Earth, delays due to signal reflections by tall
buildings, and satellites’ locations misreporting errors.
◮ Differential GPS, or DGPS, uses a stationary
reference receiver at a location with known coordinates. ◭
By calculating the difference between its location on the
basis of the GPS estimate and its true location, the
reference receiver establishes coordinate correction
factors and transmits them to all DGPS receivers in the
area. Application of the correction information usually
reduces the location inaccuracy down to about 1 m.
3-5 DIVERGENCE OF A VECTOR FIELD
153
Exercise 3-11: The power density radiated by a star
(Fig. E3.11(a)) decreases radially as S(R) = S0 /R2 , where
R is the radial distance from the star and S0 is a constant.
Recalling that the gradient of a scalar function denotes the
maximum rate of change of that function per unit distance
and the direction of the gradient is along the direction of
maximum increase, generate an arrow representation of
∇S.
where x is measured in kilometers and x = 0 is the sea–
land boundary. (a) In which direction does ∇T point and
(b) at what value of x is it a maximum?
T
T2
T1
S
Sea
Land
x
(a)
∆
T
(a)
Sea
Land
x
(b)
∆
S
Figure E3.12
Answer: (a) +x̂; (b) at x = 0.
T2 − T1
,
e−x + 1
∂T
e−x (T2 − T1 )
∇T = x̂
.
= x̂
∂x
(e−x + 1)2
T (x) = T1 +
(b)
(See
EM
.)
Figure E3.11
Answer: ∇S = −R̂ 2S0 /R3 (Fig. E3.11(b)). (See
EM
.)
Exercise 3-12: The graph in Fig. E3.12(a) depicts a gen-
tle change in atmospheric temperature from T1 over the
sea to T2 over land. The temperature profile is described
by the function
T (x) = T1 +
T2 − T1
,
e−x + 1
3-5 Divergence of a Vector Field
From our brief introduction of Coulomb’s law in Chapter 1, we
know that an isolated, positive point charge q induces an electric field E in the space around it with the direction of E being
outward away from the charge. Also, the strength (magnitude)
of E is proportional to q and decreases with distance R from
the charge as 1/R2 . In a graphical presentation, a vector field is
usually represented by field lines, as shown in Fig. 3-21. The
arrowhead denotes the direction of the field at the point where
the field line is drawn, and the length of the line provides a
qualitative depiction of the field’s magnitude.
154
CHAPTER 3 VECTOR ANALYSIS
E
n̂4
E
(x, y + Δy, z)
n̂
Δx
Face 4
Δz
+q
E
Face 2
Face 1
n̂1
Imaginary
spherical
surface
E
n̂2
Δy
(x, y, z)
Face 3
(x + Δx, y, z)
y
(x, y, z + Δz)
x
nˆ 3
Figure 3-21 Flux lines of the electric field E due to a positive
z
charge q.
Figure 3-22
At a surface boundary, flux density is defined as the amount
of outward flux crossing a unit surface ds:
Flux density of E =
E · ds E · n̂ ds
=
= E · n̂,
|ds|
ds
(3.85)
where n̂ is the normal to ds. The total flux outwardly crossing a
closed surface S, such as the enclosed surface of the imaginary
sphere outlined in Fig. 3-21, is
Total flux =
Z
S
E · ds.
(3.86)
Let us now consider the case of a differential rectangular
parallelepiped, such as a cube, whose edges align with the
Cartesian axes shown in Fig. 3-22. The edges are of lengths
∆x along x, ∆y along y, and ∆z along z. A vector field E(x, y, z)
exists in the region of space containing the parallelepiped, and
we wish to determine the flux of E through its total surface S.
Since S includes six faces, we need to sum up the fluxes
through all of them, and by definition, the flux through any
face is the outward flux from the volume ∆υ through that face.
Let E be defined as
E = x̂Ex + ŷEy + ẑEz .
(3.87)
The area of the face marked 1 in Fig. 3-22 is ∆y ∆z, and its unit
vector n̂1 = −x̂. Hence, the outward flux F1 through face 1 is
F1 =
=
Z
Z
Face 1
E · n̂1 ds
Face 1
(x̂Ex + ŷEy + ẑEz ) · (−x̂) dy dz
≈ −Ex (1) ∆y ∆z,
Flux lines of a vector field E passing
through a differential rectangular parallelepiped of volume
∆υ = ∆x ∆y ∆z.
where Ex (1) is the value of Ex at the center of face 1.
Approximating Ex over face 1 by its value at the center is
justified by the assumption that the differential volume under
consideration is very small.
Similarly, the flux out of face 2 (with n̂2 = x̂) is
F2 = Ex (2) ∆y ∆z,
where Ex (2) is the value of Ex at the center of face 2. Over a
differential separation ∆x between the centers of faces 1 and 2,
Ex (2) is related to Ex (1) by
Ex (2) = Ex (1) +
∂ Ex
∆x,
∂x
(3.90)
where we have ignored higher-order terms involving (∆x)2 and
higher powers because their contributions are negligibly small
when ∆x is very small. Substituting Eq. (3.90) into Eq. (3.89)
gives
∂ Ex
∆x ∆y ∆z.
(3.91)
F2 = Ex (1) +
∂x
The sum of the fluxes out of faces 1 and 2 is obtained by adding
Eqs. (3.88) and (3.91),
F1 + F2 =
(3.88)
(3.89)
∂ Ex
∆x ∆y ∆z.
∂x
(3.92a)
3-5 DIVERGENCE OF A VECTOR FIELD
155
Repeating the same procedure to each of the other face pairs
leads to
F3 + F4 =
∂ Ey
∆x ∆y ∆z,
∂y
(3.92b)
F5 + F6 =
∂ Ez
∆x ∆y ∆z.
∂z
(3.92c)
The sum of fluxes F1 through F6 gives the total flux through
surface S of the parallelepiped:
Z
∂ Ex ∂ Ey ∂ Ez
∆x ∆y ∆z = (div E) ∆υ ,
+
+
E · ds =
∂x
∂y
∂z
S
(3.93)
where ∆υ = ∆x ∆y ∆z and div E is a scalar function called the
divergence of E, specified in Cartesian coordinates as
div E =
∂ Ex ∂ Ey ∂ Ez
+
+
.
∂x
∂y
∂z
The divergence is a differential operator, it always operates on
vectors, and the result of its operation is a scalar. This is in
contrast with the gradient operator, which always operates on
scalars and results in a vector. Expressions for the divergence
of a vector in cylindrical and spherical coordinates are provided in Appendix C.
The divergence operator is distributive. That is, for any pair
of vectors E1 and E2 ,
∇ ·(E1 + E2 ) = ∇ · E1 + ∇ · E2 .
If ∇ · E = 0, the vector field E is called divergenceless.
The result given by Eq. (3.93) for a differential volume ∆υ
can be extended to relate the volume integral of ∇ · E over any
volume υ to the flux of E through the closed surface S that
bounds υ . That is,
(3.94)
Z
◮ By shrinking the volume ∆υ to zero, we define the
divergence of E at a point as the net outward flux per unit
volume over a closed incremental surface. ◭
∆υ →0
Z
S
E · ds
∆υ
,
∇ · E dυ =
S
E · ds.
(3.98)
(divergence theorem)
(3.95)
Example 3-13: Calculating the Divergence
where S encloses the elemental volume ∆υ . Instead of denoting the divergence of E by div E, it is common practice to
denote it as ∇ · E. That is,
∇ · E = div E =
υ
Z
This relationship, known as the divergence theorem, is used
extensively in electromagnetics.
Thus, from Eq. (3.93), we have
div E , lim
(3.97)
∂ Ex ∂ Ey ∂ Ez
+
+
∂x
∂y
∂z
(3.96)
for a vector E in Cartesian coordinates.
◮ From the definition of the divergence of E given by
Eq. (3.95), field E has positive divergence if the net flux
out of surface S is positive, which may be “viewed” as
if volume ∆υ contains a source of field lines. If the
divergence is negative, ∆υ may be viewed as containing
a sink of field lines because the net flux is into ∆υ . For
a uniform field E, the same amount of flux enters ∆υ as
leaves it; hence, its divergence is zero and the field is said
to be divergenceless. ◭
Determine the divergence of each of the following vector fields
and then evaluate them at the indicated points:
(a) E = x̂3x2 + ŷ2z + ẑx2 z at (2, −2, 0);
(b) E = R̂(a3 cos θ /R2 ) − θ̂θ(a3 sin θ /R2 ) at (a/2, 0, π ).
Solution: (a)
∇·E =
=
∂ Ex ∂ Ey ∂ Ez
+
+
∂x
∂y
∂z
∂
∂
∂
(3x2 ) + (2z) + (x2 z)
∂x
∂y
∂z
= 6x + 0 + x2 = x2 + 6x.
At (2, −2, 0), ∇ · E
(2,−2,0)
= 16.
156
CHAPTER 3 VECTOR ANALYSIS
Module 3.3 Divergence Select a vector function f(x, y, z), evaluate its divergence, and display both in an appropriate 2-D
plane.
(b) From the expression given in Appendix C for the divergence of a vector in spherical coordinates, it follows that
∂
1
1 ∂
(Eθ sin θ )
(R2 ER ) +
2
R ∂R
R sin θ ∂ θ
1 ∂ Eφ
+
R sin θ ∂ φ
3 2 1 ∂ 3
1
∂
a sin θ
= 2
(a cos θ ) +
−
R ∂R
R sin θ ∂ θ
R2
∇·E =
= 0−
2a3 cos θ
2a3 cos θ
=−
.
3
R
R3
Answer: ∇ · A = 0. (See
∇ · A at (2, 0, 3).
Answer: ∇ · A = 6. (See
= −16.
EM
.)
Exercise 3-15: If E = R̂AR in spherical coordinates,
calculate the flux of E through a spherical surface of
radius a, centered at the origin.
Z
S
(a/2,0,π )
.)
φr sin φ + ẑ3z, find
Exercise 3-14: Given A = r̂r cos φ + φ̂
Answer:
At R = a/2 and θ = 0, ∇ · E
EM
E · ds = 4π Aa3. (See
EM
.)
Exercise 3-16: Verify the divergence theorem by calculatExercise 3-13: Given A = e−2y (x̂ sin 2x + ŷ cos 2x), find
∇ · A.
ing the volume integral of the divergence of the field E of
Exercise 3.15 over the volume bounded by the surface of
radius a.
3-6 CURL OF A VECTOR FIELD
157
Exercise 3-17: The arrow representation in Fig. E3.17
represents the vector field A = x̂ x − ŷ y. At a given point
in space, A has a positive divergence ∇ · A if the net
flux flowing outward through the surface of an imaginary
infinitesimal volume centered at that point is positive,
∇ · A is negative if the net flux is into the volume, and
∇·A = 0 if the same amount of flux enters into the volume
as leaves it. Determine ∇ · A everywhere in the x–y plane.
y
a
d
Contour C
Δx
Δx
c
b
B
x
(a) Uniform field
z
Current I
Figure E3.17
φ̂
Answer: ∇ · A = 0 everywhere. (See
EM
Contour C
.)
φ
3-6 Curl of a Vector Field
B
So far we have defined and discussed two of the three fundamental operators used in vector analysis: the gradient of
a scalar and the divergence of a vector. Now we introduce
the curl operator. The curl of a vector field B describes its
rotational property, or circulation. The circulation of B is
defined as the line integral of B around a closed contour C;
Circulation =
Z
C
B · dl.
(3.99)
To gain a physical understanding of this definition, we consider
two examples. The first is for a uniform field B = x̂B0 , whose
field lines are as depicted in Fig. 3-23(a). For the rectangular
contour abcd shown in the figure, we have
Circulation =
Z b
a
+
x̂B0 · x̂ dx +
Z d
c
Z c
b
x̂B0 · x̂ dx +
y
r
x̂B0 · ŷ dy
Z a
= B0 ∆x − B0 ∆x = 0,
d
x̂B0 · ŷ dy
(3.100)
where ∆x = b − a = c − d and, because x̂ · ŷ = 0, the second
and fourth integrals are zero. According to Eq. (3.100), the
circulation of a uniform field is zero.
x
(b) Azimuthal field
Figure 3-23 Circulation is zero for the uniform field in (a), but
it is not zero for the azimuthal field in (b).
Next, we consider the magnetic flux density B induced by
an infinite wire carrying a dc current I. If the current is in
free space and it is oriented along the z direction, then from
Eq. (1.13),
µ0 I
B = φ̂φ
,
(3.101)
2π r
where µ0 is the permeability of free space and r is the radial
distance from the current in the x–y plane. The direction of B
is along the azimuth unit vector φ̂φ. The field lines of B are
concentric circles around the current, as shown in Fig. 3-23(b).
158
CHAPTER 3 VECTOR ANALYSIS
For a circular contour C of radius r centered at the origin in the
x–y plane, the differential length vector dl = φ̂φr d φ , and the
circulation of B is
nˆ
Circulation =
Z
C
B · dl =
Z 2π
φ̂φ
0
ds = n
ˆ ds
µ0 I
· φ̂φr d φ = µ0 I. (3.102)
2π r
In this case, the circulation is not zero. However, had the
contour C been in the x–z or y–z planes, dl would not have
had a φ̂φ component, and the integral would have yielded a
zero circulation. Clearly, the circulation of B depends on the
choice of contour and the direction in which it is traversed. To
describe the circulation of a tornado, for example, we would
like to choose our contour such that the circulation of the
wind field is maximum, and we would like the circulation
to have both a magnitude and a direction with the direction
being toward the tornado’s vortex. The curl operator embodies
these properties. The curl of a vector field B, denoted curl B or
× B, is defined as
∇×
Z
1
n̂ B · dl
.
∆s→0 ∆s
C
max
× B = curl B = lim
∇×
ds
S
dl
contour C
Figure 3-24 The direction of the unit vector n̂ is along the
thumb when the other four fingers of the right hand follow dl.
3-6.1
For any two vectors A and B and scalar V ,
(3.103)
◮ Curl B is the circulation of B per unit area, with the
area ∆s of the contour C being oriented such that the
circulation is maximum. ◭
The direction of curl B is n̂, the unit normal of ∆s, defined
according to the right-hand rule with the four fingers of the
right hand following the contour direction dl and the thumb
×B
pointing along n̂ (Fig. 3-24). When we use the notation ∇×
to denote curl B, it should not be interpreted as the cross
product of ∇ and B.
For a vector B specified in Cartesian coordinates as
B = x̂Bx + ŷBy + ẑBz ,
=
ŷ
∂
∂y
By
ẑ
∂
∂z
Bz
.
3-6.2
× (A + B) = ∇×
× A + ∇×
× B,
(1) ∇×
(3.106a)
× A) = 0,
(2) ∇ ·(∇×
(3.106b)
× (∇V ) = 0.
(3) ∇×
(3.106c)
Stokes’s Theorem
◮ Stokes’s theorem converts the surface integral of the
curl of a vector over an open surface S into a line
integral of the vector along the contour C bounding the
surface S. ◭
(3.104)
it can be shown, through a rather lengthy derivation, that
Eq. (3.103) leads to
∂ Bz ∂ By
∂ Bx ∂ Bz
× B = x̂
∇×
+ ŷ
−
−
∂y
∂z
∂z
∂x
∂ By ∂ Bx
−
+ ẑ
∂x
∂y
x̂
∂
∂x
Bx
Vector Identities Involving the Curl
(3.105)
Expressions for ∇ × B are given in Appendix C for the three
orthogonal coordinate systems considered in this chapter.
For the geometry shown in Fig. 3-24, Stokes’s theorem states
Z
S
× B) · ds =
(∇×
Z
C
B · dl.
(3.107)
(Stokes’s theorem)
Its validity follows from the definition of ∇ × B given by
× B = 0, the field B is said to be conservative
Eq. (3.103). If ∇×
or irrotational because its circulation, represented by the righthand side of Eq. (3.107), is zero, irrespective of the contour
chosen.
3-6 CURL OF A VECTOR FIELD
Technology Brief 6:
X-Ray Computed Tomography
◮ The word tomography is derived from the Greek
words tome, meaning section or slice, and graphia,
meaning writing. ◭
Computed tomography, also known as CT scan or
CAT scan (for computed axial tomography), refers to
a technique capable of generating 3-D images of the
X-ray attenuation (absorption) properties of an object.
This is in contrast to the traditional, X-ray technique that
produces only a 2-D profile of the object (Fig. TF6-1).
CT was invented in 1972 by British electrical engineer
Godfrey Hounsfeld and independently by Allan Cormack , a South African-born American physicist. The two
inventors shared the 1979 Nobel Prize in Physiology or
Figure TF6-1 2-D X-ray image.
159
Medicine. Among diagnostic imaging techniques, CT has
the decided advantage in having the sensitivity to image
body parts on a wide range of densities—from soft tissue
to blood vessels and bones.
Principle of Operation
In the system shown in Fig. TF6-2, the X-ray source and
detector array are contained inside a circular structure
through which the patient is moved along a conveyor belt.
A CAT scan technician can monitor the reconstructed
images to ensure that they do not contain artifacts such
as streaks or blurry sections caused by movement on the
part of the patient during the measurement process.
A CT scanner uses an X-ray source with a narrow
slit that generates a fan-beam that is wide enough to
encompass the extent of the body but only a few millimeters in thickness (Fig. TF6-3(a)). Instead of recording
the attenuated X-ray beam on film, it is captured by an
array of some 700 detectors. The X-ray source and
the detector array are mounted on a circular frame that
rotates in steps of a fraction of a degree over a full 360◦
circle around the patient, each time recording an X-ray
attenuation profile from a different angular perspective.
Typically, 1,000 such profiles are recorded per each thin
traverse slice of anatomy. In today’s technology, this
process is completed in less than 1 second. To image
an entire part of the body, such as the chest or head, the
process is repeated over multiple slices (layers), which
typically takes about 10 seconds to complete.
Figure TF6-2 CT scanner.
160
CHAPTER 3 VECTOR ANALYSIS
Image Reconstruction
X-ray
source
Fan beam
of X-rays
Detector
array
For each anatomical slice, the CT scanner generates
on the order of 7 × 105 measurements (1,000 angular
orientations × 700 detector channels). Each measurement represents the integrated path attenuation for the
narrow beam between the X-ray source and the detector
(Fig. TF6-3(b)), and each volume element (voxel) contributes to 1,000 such measurement beams.
◮ Commercial CT machines use a technique called
filtered back-projection to “reconstruct” an image
of the attenuation rate of each voxel in the anatomical slice and, by extension, for each voxel in the
entire body organ. This is accomplished through
the application of a sophisticated matrix inversion
process. ◭
Computer
and monitor
(a) CT scanner
X-ray source
Voxel
Detector
(b) Detector measures integrated attenuation
along anatomical path
(c) CT image of a normal brain
Figure TF6-3 Basic elements of a CT scanner.
A sample CT image of the brain is shown in
Fig. TF6-3(c).
3-6 CURL OF A VECTOR FIELD
Example 3-14:
161
× B over the specified surface S is
The integral of ∇×
Verification of Stokes’s
Theorem
Z
S
For vector field B = ẑ cos φ /r, verify Stokes’s theorem
for a segment of a cylindrical surface defined by r = 2,
π /3 ≤ φ ≤ π /2, and 0 ≤ z ≤ 3 (Fig. 3-25).
z
z=3
r=2
× B) · ds =
(∇×
=
Z 3 Z π /2 z=0 φ =π /3
Z 3 Z π /2
π /3
0
Z
C
c
B · dl =
Z b
a
b
π/3
· r̂r d φ dz
Z d
Z c
b
Bcd · dl +
Bbc · dl
Z a
d
Bda · dl,
where Bab , Bbc , Bcd , and Bda are the field B along segments ab, bc, cd, and da, respectively. Over segment ab,
the dot product of Bab = ẑ (cos φ ) /2 and dl = φ̂φr d φ is
zero, and the same is true for segment cd. Over segment bc,
φ = π /2; hence, Bbc = ẑ(cos π /2)/2 = 0. For the last segment,
Bda = ẑ(cos π /3)/2 = ẑ/4 and dl = ẑ dz. Hence,
nˆ = rˆ
π/2
3
3
sin φ
d φ dz = − = − .
r
2r
4
Bab · dl +
c
d
0
sin φ
cos φ
+ φ̂φ 2
r2
r
Right-hand side: The surface S is bounded by contour
C = abcd shown in Fig. 3-25. The direction of C is chosen
so that it is compatible with the surface normal r̂ by the righthand rule. Hence,
+
2
−
−r̂
y
a
x
Z
C
Figure 3-25 Geometry of Example 3-14.
B · dl =
Z a 1
ẑ
d
4
· ẑ dz =
Z 0
1
3
3
dz = − ,
4
4
which is the same as the result obtained by evaluating the
left-hand side of Stokes’s equation.
Solution: Stokes’s theorem states that
Z
S
× B) · ds =
(∇×
Z
C
Exercise 3-18: Find ∇ × A at (2, 0, 3) in cylindrical
coordinates for the vector field
B · dl.
A = r̂10e−2r cos φ + ẑ10 sin φ .
Left-hand side: With B having only a component Bz = cos φ /r,
use of the expression for ∇ × B in cylindrical coordinates in
Appendix C gives
1 ∂ Bz ∂ Bφ
∂ Br ∂ Bz
+ φ̂φ
−
−
r ∂φ
∂z
∂z
∂r
1 ∂
∂ Br
(rBφ ) −
+ ẑ
r ∂r
∂φ
cos φ
1 ∂
∂ cos φ
− φ̂φ
= r̂
r ∂φ
r
∂r
r
× B = r̂
∇×
sin φ
cos φ
= −r̂ 2 + φ̂φ 2 .
r
r
Answer: (See
×A =
∇×
r̂
EM
.)
10 cos φ ẑ 10e−2r
+
sin φ
r
r
= r̂5.
(2,0,3)
Exercise 3-19: Find ∇ × A at (3, π /6, 0) in spherical
coordinates for the vector field A = θ̂θ12 sin θ .
Answer: (See
EM
.)
× A = φ̂φ
∇×
12 sin θ
R
= φ̂φ2.
(3,π /6,0)
162
CHAPTER 3 VECTOR ANALYSIS
Module 3.4 Curl Select a vector f(x, y), evaluate its curl, and display both in the x–y plane.
3-7 Laplacian Operator
That is,
In later chapters, we sometimes deal with problems involving
multiple combinations of operations on scalars and vectors. A
frequently encountered combination is the divergence of the
gradient of a scalar. For a scalar function V defined in Cartesian
coordinates, its gradient is
∂V
∂V
∂V
+ ŷ
+ ẑ
= x̂Ax + ŷAy + ẑAz = A,
∂x
∂y
∂z
(3.108)
where we defined a vector A with components Ax = ∂ V /∂ x,
Ay = ∂ V /∂ y, and Az = ∂ V /∂ z. The divergence of ∇V is
∇V = x̂
∇ ·(∇V ) = ∇ · A =
=
∂ x2
+
∂ 2V
∂ y2
+
∂ 2V
∂ z2
∂ 2V ∂ 2V ∂ 2V
+ 2 + 2 .
∂ x2
∂y
∂z
(3.110)
As we can see from Eq. (3.110), the Laplacian of a scalar
function is a scalar. Expressions for ∇2V in cylindrical and
spherical coordinates are given in Appendix C.
The Laplacian of a scalar can be used to define the Laplacian
of a vector. For a vector E specified in Cartesian coordinates
as
E = x̂Ex + ŷEy + ẑEz ,
(3.111)
the Laplacian of E is
∂ Ax ∂ Ay ∂ Az
+
+
∂x
∂y
∂z
∂ 2V
∇2V = ∇ · (∇V ) =
2
.
(3.109)
For convenience, ∇ ·(∇V ) is called the Laplacian of V and is
denoted by ∇2V (the symbol ∇2 is pronounced “del square”).
∇ E=
∂2
∂2
∂2
+
+
∂ x2 ∂ y2 ∂ z2
E = x̂ ∇2 Ex + ŷ ∇2 Ey + ẑ∇2 Ez .
(3.112)
Thus, in Cartesian coordinates the Laplacian of a vector is
a vector whose components are equal to the Laplacians of
the vector components. Through direct substitution, it can be
3-7 LAPLACIAN OPERATOR
163
shown that
If a vector field is solenoidal
at a given point in space, does it necessarily follow that
the vector field is zero at that point? Explain.
Concept Question 3-14:
× (∇×
× E).
∇2 E = ∇(∇ · E) − ∇×
(3.113)
What is the meaning of the
transformation provided by the divergence theorem?
Concept Question 3-15:
What do the magnitude and
direction of the gradient of a scalar quantity represent?
Concept Question 3-11:
How is the curl of a vector
field at a point related to the circulation of the vector field?
Concept Question 3-16:
Prove the
Eq. (3.84c) in Cartesian coordinates.
Concept Question 3-12:
validity
of
What is the meaning of the
transformation provided by Stokes’s theorem?
Concept Question 3-17:
What is the physical meaning
of the divergence of a vector field?
Concept Question 3-13:
Concept Question 3-18:
When is a vector field “con-
servative”?
Chapter 3 Summary
Concepts
• Vector algebra governs the laws of addition, subtraction, and multiplication of vectors, and vector calculus
encompasses the laws of differentiation and integration
of vectors.
• In a right-handed orthogonal coordinate system, the
three base vectors are mutually perpendicular to each
other at any point in space, and the cyclic relations
governing the cross products of the base vectors obey
the right-hand rule.
• The dot product of two vectors produces a scalar,
whereas the cross product of two vectors produces
another vector.
• A vector expressed in a given coordinate system can be
expressed in another coordinate system through the use
of transformation relations linking the two coordinate
systems.
• The fundamental differential functions in vector calculus are the gradient, the divergence, and the curl.
• The gradient of a scalar function is a vector whose
magnitude is equal to the maximum rate of increasing
•
•
•
•
•
change of the scalar function per unit distance, and its
direction is along the direction of maximum increase.
The divergence of a vector field is a measure of the net
outward flux per unit volume through a closed surface
surrounding the unit volume.
The divergence theorem transforms the volume integral
of the divergence of a vector field into a surface integral
of the field’s flux through a closed surface surrounding
the volume.
The curl of a vector field is a measure of the circulation
of the vector field per unit area ∆s, with the orientation
of ∆s chosen such that the circulation is maximum.
Stokes’s theorem transforms the surface integral of the
curl of a vector field into a line integral of the field over
a contour that bounds the surface.
The Laplacian of a scalar function is defined as the
divergence of the gradient of that function.
164
CHAPTER 3 VECTOR ANALYSIS
Mathematical and Physical Models
Distance Between Two Points
Vector Operators
d = [(x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 ]1/2
1/2
d = r22 +r12 − 2r1 r2 cos(φ2 −φ1 )+(z2 −z1 )2
d = R22 + R21 − 2R1R2 [cos θ2 cos θ1
1/2
+ sin θ1 sin θ2 cos(φ2 − φ1 )]
Coordinate Systems
Table 3-1
Coordinate Transformations
Table 3-2
Vector Products
∇T = x̂
∂ Ex ∂ Ey ∂ Ez
+
+
∂x
∂y
∂z
∂ Bz ∂ By
∂ Bx ∂ Bz
× B = x̂
+ ŷ
∇×
−
−
∂y
∂z
∂z
∂x
∂ By ∂ Bx
+ ẑ
−
∂x
∂y
∇·E =
A · B = AB cos θAB
∇2V =
× B = n̂ AB sin θAB
A×
× A) = C ·(A×
× B)
A ·(B × C) = B ·(C×
∂T
∂T
∂T
+ ŷ
+ ẑ
∂x
∂y
∂z
∂ 2V ∂ 2V ∂ 2V
+ 2 + 2
∂ x2
∂y
∂z
(see Appendix C for cylindrical
and spherical coordinates)
× (B×
× C) = B(A · C) − C(A · B)
A×
Divergence Theorem
Stokes’s Theorem
Z
Z
υ
∇ · E dυ =
Z
S
E · ds
Important Terms
S
× B) · ds =
(∇×
Z
C
B · dl
Provide definitions or explain the meaning of the following terms:
azimuth angle
base vectors
Cartesian coordinate system
circulation of a vector
conservative field
cross product
curl operator
cylindrical coordinate system
differential area vector
differential length vector
differential volume
directional derivative
distance vector
divergence operator
divergence theorem
divergenceless
dot product
field lines
flux density
flux lines
gradient operator
irrotational field
Laplacian operator
magnitude
orthogonal coordinate system
position vector
PROBLEMS
Section 3-1: Vector Algebra
∗ 3.1 Vector A starts at point (1, −1, −3) and ends at point
(2, −1, 0). Find a unit vector in the direction of A.
∗ Answer(s) available in Appendix E.
radial distance r
range R
right-hand rule
scalar product
scalar quantity
simple product
solenoidal field
spherical coordinate system
Stokes’s theorem
vector product
vector quantity
unit vector
zenith angle
3.2 Given vectors A = x̂2 − ŷ3 + ẑ, B = x̂2 − ŷ + ẑ3, and
C = x̂4 + ŷ2 − ẑ2, show that C is perpendicular to both A
and B.
∗ 3.3 In Cartesian coordinates, the three corners of a triangle
are P1 = (0, 4, 4), P2 = (4, −4, 4), and P3 = (2, 2, −4). Find the
area of the triangle.
PROBLEMS
165
3.4 Given A = x̂2 − ŷ3 + ẑ1 and B = x̂Bx + ŷ2 + ẑBz :
(a) Find Bx and Bz if A is parallel to B.
(b) Find a relation between Bx and Bz if A is perpendicular
to B.
3.5 Given vectors A = x̂ + ŷ2 − ẑ3,
C = ŷ2 − ẑ4, find the following:
B = x̂2 − ŷ4, and
∗(a) A and â
(b) The component of B along C
(c) θAC
×C
(d) A×
∗(e) A ·(B×
× C)
× (B×
× C)
(f) A×
×B
(g) x̂×
Vector A starts at the origin and ends at point P on the line such
that A is orthogonal to the line. Find an expression for A.
3.14 Show the following given two vectors A and B:
(a) The vector C defined as the vector component of B in the
direction of A is given by
C = â(B · â) =
where â is the unit vector of A.
(b) The vector D defined as the vector component of B
perpendicular to A is given by
D = B−
∗(h) (A×
× ŷ) · ẑ
3.6 Given vectors A = x̂2 − ŷ + ẑ3 and B = x̂3 − ẑ2, find
a vector C whose magnitude is 9 and whose direction is
perpendicular to both A and B.
3.7 Given A = x̂(x + 2y) − ŷ(y + 3z) + ẑ(3x − y), determine
a unit vector parallel to A at point P = (1, −1, 2).
3.8 By expansion in Cartesian coordinates, prove:
(a) The relation for the scalar triple product given by
Eq. (3.29).
(b) The relation for the vector triple product given by
Eq. (3.33).
∗ 3.9 Find an expression for the unit vector directed toward the
origin from an arbitrary point on the line described by x = 1
and z = −2.
3.10 Find an expression for the unit vector directed toward
the point P located on the z axis at a height h above the
x–y plane from an arbitrary point Q = (x, y, −5) in the plane
z = −5.
∗ 3.11 Find a unit vector parallel to either direction of the line
described by
2x + z = 4
3.12 Two lines in the x–y plane are described by the following expressions:
Line 1
Line 2
x + 2y = −6
3x + 4y = 8
Use vector algebra to find the smaller angle between the lines
at their intersection point.
∗ 3.13 A given line is described by
x + 2y = 4.
A(B · A)
,
|A|2
A(B · A)
|A|2
∗ 3.15 A certain plane is described by
2x + 3y + 4z = 16
Find the unit vector normal to the surface in the direction away
from the origin.
3.16 Given B = x̂(z − 3y) + ŷ(2x − 3z) − ẑ(x + y), find a unit
vector parallel to B at point P = (1, 0, −1).
∗ 3.17 Find a vector G whose magnitude is 4 and whose
direction is perpendicular to both vectors E and F, where
E = x̂ + ŷ2 − ẑ2 and F = ŷ 3 − ẑ6.
3.18 A given line is described by the equation:
y = x − 1.
Vector A starts at point P1 = (0, 2) and ends at point P2 on the
line, at which A is orthogonal to the line. Find an expression
for A.
3.19 Vector field E is given by
E = R̂ 5R cos θ − θ̂θ
12
sin θ cos φ + φ̂φ3 sin φ .
R
Determine the component of E tangential to the spherical
surface R = 2 at point P = (2, 30◦, 60◦ ).
3.20 When sketching or demonstrating the spatial variation
of a vector field, we often use arrows, as in Fig. P3.20, wherein
the length of the arrow is made to be proportional to the
strength of the field and the direction of the arrow is the same
as that of the field’s. The sketch shown in Fig. P3.20, which
represents the vector field E = r̂r, consists of arrows pointing
radially away from the origin and their lengths increasing
linearly in proportion to their distance away from the origin.
166
CHAPTER 3 VECTOR ANALYSIS
Using this arrow representation, sketch each of the following
vector fields:
(a) E1 = −x̂y
(b) E2 = ŷx
(c) E3 = x̂x + ŷy
(d) E4 = x̂x + ŷ2y
(e) E5 = φ̂φr
(f) E6 = r̂ sin φ
∗ (d) P = (−2, 2, −2)
4
3.23 Convert the coordinates of the following points from
cylindrical to Cartesian coordinates:
(a) P1 = (2, π /4, −3)
(b) P2 = (3, 0, −2)
(c) P3 = (4, π , 5)
3.24 Convert the coordinates of the following points from
spherical to cylindrical coordinates:
∗ (a) P = (5, 0, 0)
1
(b) P2 = (5, 0, π )
(c) P3 = (3, π /2, 0)
y
E
3.25 Use the appropriate expression for the differential surface area ds to determine the area of each of the following
surfaces:
(a) r = 3; 0 ≤ φ ≤ π /3; −2 ≤ z ≤ 2
(b) 2 ≤ r ≤ 5; π /2 ≤ φ ≤ π ; z = 0
E
∗ (c) 2 ≤ r ≤ 5; φ = π /4; −2 ≤ z ≤ 2
x
(d) R = 2; 0 ≤ θ ≤ π /3; 0 ≤ φ ≤ π
(e) 0 ≤ R ≤ 5; θ = π /3; 0 ≤ φ ≤ 2π
Also sketch the outline of each surface.
3.26 Find the volumes described by the following:
E
E
Figure P3.20 Arrow representation for vector field E = r̂ r
(Problem 3.20).
3.21 Use arrows to sketch each of the following vector fields:
(a) E1 = x̂x − ŷy
(b) E2 = −φ̂φ
(c) E3 = ŷ (1/x)
(d) E4 = r̂ cos φ
Sections 3-2 and 3-3: Coordinate Systems
3.22 Convert the coordinates of the following points from
Cartesian to cylindrical and spherical coordinates:
∗ (a) P = (1, 2, 0)
1
(b) P2 = (0, 0, 2)
(c) P3 = (1, 1, 3)
∗ (a) 2 ≤ r ≤ 5; π /2 ≤ φ ≤ π ; 0 ≤ z ≤ 2
(b) 0 ≤ R ≤ 5; 0 ≤ θ ≤ π /3; 0 ≤ φ ≤ 2π
Also sketch the outline of each volume.
3.27 A section of a sphere is described by 0 ≤ R ≤ 2,
0 ≤ θ ≤ 90◦ , and 30◦ ≤ φ ≤ 90◦ . Find the following:
(a) The surface area of the spherical section.
(b) The enclosed volume.
Also sketch the outline of the section.
3.28 A vector field is given in cylindrical coordinates by
E = r̂r cos φ + φ̂φr sin φ + ẑz2 .
Point P = (2, π , 3) is located on the surface of the cylinder
described by r = 2. At point P, find:
(a) The vector component of E perpendicular to the cylinder.
(b) The vector component of E tangential to the cylinder.
3.29 At a given point in space, vectors A and B are given in
spherical coordinates by
A = R̂4 + θ̂θ2 − φ̂φ,
B = −R̂2 + φ̂φ3.
PROBLEMS
Find:
(a) The scalar component, or projection, of B in the direction
of A.
(b) The vector component of B in the direction of A.
(c) The vector component of B perpendicular to A.
∗ 3.30 Given vectors
A = r̂(cos φ + 3z) − φ̂φ(2r + 4 sin φ ) + ẑ(r − 2z),
B = −r̂sin φ + ẑ cos φ ,
find
(a) θAB at (2, π /2, 0)
(b) A unit vector perpendicular to both A and B at (2, π /3, 1)
167
(b) B = ŷ(x2 + y2 + z2 ) − ẑ(x2 + y2 ) at P2 = (−1, 0, 2)
∗ (c) C = r̂ cos φ − φ̂φ sin φ + ẑcos φ sin φ at
P3 = (2, π /4, 2)
(d) D = x̂y2 /(x2 + y2 ) − ŷx2 /(x2 + y2) + ẑ4 at
P4 = (1, −1, 2)
Sections 3-4 to 3-7: Gradient, Divergence,
and Curl Operators
3.36 Find the gradient of the following scalar functions:
(a) T = 3/(x2 + z2 )
(b) V = xy2 z4
(c) U = z cos φ /(1 + r2)
(d) W = e−R sin θ
3.31 Find the distance between the following pairs of points:
(a) P1 = (1, 2, 3) and P2 = (−2, −3, −2) in Cartesian coordinates.
(b) P3 = (1, π /4, 3) and P4 = (3, π /4, 4) in cylindrical coordinates.
(c) P5 = (4, π /2, 0) and P6 = (3, π , 0) in spherical coordinates.
∗ (e) S = 4x2 e−z + y3
3.32 Determine the distance between the following pairs of
points:
∗ (b) T = x2 , for −10 ≤ x ≤ 10
∗ (a) P = (1, 1, 2) and P = (0, 2, 3)
1
2
(b) P3 = (2, π /3, 1) and P4 = (4, π /2, 3)
(c) P5 = (3, π , π /2) and P6 = (4, π /2, π )
3.33 Transform the vector
A = R̂ sin2 θ cos φ + θ̂θ cos2 φ − φ̂φ sin φ
into cylindrical coordinates and then evaluate it at
P = (2, π /2, π /2).
3.34 Transform the following vectors into cylindrical coordinates and then evaluate them at the indicated points:
(a) A = x̂(x + y) at P1 = (1, 2, 3)
(b) B = x̂(y − x) + ŷ(x − y) at P2 = (1, 0, 2)
∗(c) C = x̂y2 /(x2 + y2 ) − ŷx2 /(x2 + y2 ) + ẑ4 at
P3 = (1, −1, 2)
(d) D = R̂ sin θ + θ̂θ cos θ + φ̂φ cos2 φ at
P4 = (2, π /2, π /4)
∗(e) E = R̂ cos φ + θ̂θ sin φ + φ̂φ sin2 θ at P = (3, π /2, π )
5
3.35 Transform the following vectors into spherical coordinates and then evaluate them at the indicated points:
(a) A = x̂y2 + ŷxz + ẑ4 at P1 = (1, −1, 2)
(f) N = r2 cos2 φ
(g) M = R cos θ sin φ
3.37 For each of the following scalar fields, obtain an analytical solution for ∇T and generate a corresponding arrow
representation.
(a) T = 10 + x, for −10 ≤ x ≤ 10
(b)
(c)
(e)
(f)
T
T
T
T
= 100 + xy, for −10 ≤ x ≤ 10
= x2 y2 , for −10 ≤ x, y ≤ 10
= 20 + x + y, for −10 ≤ x, y ≤ 10
= 1 + sin(π x/3), for −10 ≤ x ≤ 10
∗ (g) T = 1 + cos(π x/3), for −10 ≤ x ≤ 10
0 ≤ r ≤ 10
0 ≤ φ ≤ 2π .
0 ≤ r ≤ 10
2
(i) T = 15 + r cos φ , for
0 ≤ φ ≤ 2π .
(h) T = 15 + r cos φ , for
∗ 3.38 The gradient of a scalar function T is given by
∇T = ẑ e−3z
If T = 10 at z = 0, find T (z).
3.39 Follow a procedure similar to that leading to Eq. (3.82)
to derive the expression given by Eq. (3.83) for ∇ in spherical
coordinates.
∗ 3.40 For the scalar function V = xy2 − z2 , determine its directional derivative along the direction of vector A = (x̂ − ŷz)
and then evaluate it at P = (1, −1, 4).
3.41 Evaluate the line integral of E = x̂ x − ŷ y along the
segment P1 to P2 of the circular path shown in Fig. P3.41.
168
CHAPTER 3 VECTOR ANALYSIS
(b) A = −x̂ sin 2y + ŷcos 2x, for −π ≤ x, y ≤ π
y
P1 = (0, 3)
x
P2 = (−3, 0)
Figure P3.41 Problem 3.41.
Figure P3.44(b)
3.42 For the scalar function T = 12 e−r/5 cos φ , determine
its directional derivative along the radial direction r̂ and then
evaluate it at P = (2, π /4, 3).
∗ 3.43 For the scalar function U =
(c) A = −x̂ xy + ŷy2 , for −10 ≤ x, y ≤ 10
1
R
sin2 θ , determine its
directional derivative along the range direction R̂ and then
evaluate it at P = (5, π /4, π /2).
3.44 Each of the following vector fields is displayed in
Fig. P3.44 in the form of a vector representation. Determine
∇ · A analytically and then compare the result with your
expectations on the basis of the displayed pattern.
(a) A = −x̂ cos x sin y + ŷsin x cos y, for −π ≤ x, y ≤ π
Figure P3.44(c)
(d) A = −x̂ cos x + ŷsin y, for −π ≤ x, y ≤ π
Figure P3.44(a)
Figure P3.44(d)
PROBLEMS
169
(e) A = x̂ x, for −10 ≤ x ≤ 10
Figure P3.44(e)
(f) A = x̂ xy2 , for −10 ≤ x, y ≤ 10
Figure P3.44(f)
(g) A = x̂ xy2 + ŷ x2 y, for −10 ≤ x, y ≤ 10
Figure P3.44(g)
(h) A = x̂ sin
πx
10
+ ŷ sin
πy 10 ,
for −10 ≤ x, y ≤ 10
Figure P3.44(h)
0 ≤ r ≤ 10
(i) A = r̂ r + φ̂φ r cos φ , for
0 ≤ φ ≤ 2π .
Figure P3.44(i)
0 ≤ r ≤ 10
(j) A = r̂ r2 + φ̂φ r2 sin φ , for
0 ≤ φ ≤ 2π .
Figure P3.44(j)
170
CHAPTER 3 VECTOR ANALYSIS
3.45 The scalar function V is given by
y
y
V = 2 sin x + xy2 + ye2z.
1
1
Determine ∇V and then evaluate it at P(0, 1, 2).
3.46 The scalar function V is given by
0
(a) Determine ∇V in Cartesian coordinates.
(b) Convert the result of part (a) from Cartesian to cylindrical
coordinates.
(c) Convert the expression for V into cylindrical coordinates
and then determine ∇V in those coordinates. Compare the
results of parts (b) and (c).
∗ 3.47 Vector field E is characterized by the following properties: (a) E points along R̂; (b) the magnitude of E is a function
of only the distance from the origin; (c) E vanishes at the
origin; and (d) ∇ · E = 12, everywhere. Find an expression for
E that satisfies these properties.
3.48 For the vector field E = x̂xz − ŷyz2 − ẑxy, verify the
divergence theorem by computing
(a) the total outward flux flowing through the surface of a
cube centered at the origin and with sides equal to 2 units
each and parallel to the Cartesian axes, and
(b) the integral of ∇ · E over the cube’s volume.
x
1
2z
.
V= 2
x + y2
0
(a)
Figure P3.52 Contours for (a) Problem 3.52 and (b) Prob-
3.53 Repeat Problem 3.52 for the contour shown in
Fig. P3.52(b).
3.54 Verify Stokes’s theorem for the vector field
B = (r̂r cos φ + φ̂φ sin φ )
by evaluating
(a)
Z
C
B · dl over the semicircular contour shown in
Fig. P3.54(a), and
(b)
Z
S
× B) · ds over the surface of the semicircle.
(∇×
y
y
2
2
∗ 3.50 A vector field D = r̂r3 exists in the region between two
(a)
(b)
Z
ZS
υ
D · ds
∇ · D dυ
x
lem 3.53.
3.49 For the vector field E = r̂10e−r − ẑ3z, verify the divergence theorem for the cylindrical region enclosed by r = 2,
z = 0, and z = 4.
concentric cylindrical surfaces defined by r = 1 and r = 2 with
both cylinders extending between z = 0 and z = 5. Verify the
divergence theorem by evaluating the following:
2
1
(b)
1
–2
0
2
x
0
(a)
1
2
x
(b)
Figure P3.54 Contour paths for (a) Problem 3.54 and (b)
Problem 3.55.
3.51 For the vector field D = R̂3R2 , evaluate both sides of
the divergence theorem for the region enclosed between the
spherical shells defined by R = 1 and R = 2.
3.55 Repeat Problem 3.54 for the contour shown in
Fig. P3.54(b).
3.52 For the vector field E = x̂xy − ŷ(x2 + 2y2), calculate
3.56 Verify Stokes’s theorem for the vector field
(a)
Z
C
E · dl around the triangular contour shown in
Fig. P3.52(a).
(b)
Z
S
× E) · ds over the area of the triangle.
(∇×
A = R̂ cos θ + φ̂φ sin θ
by evaluating it on the hemisphere of unit radius.
PROBLEMS
171
3.57 Verify Stokes’s theorem for the vector field
∗ (a) A = x̂x2 − ŷ2xy
B = (r̂ cos φ + φ̂φ sin φ )
by evaluating:
(a)
Z
C
B · dℓℓ over the path comprising a quarter section of a
circle, as shown in Fig. P3.57, and
(b)
Z
S
3.58 Determine if each of the following vector fields is
solenoidal, conservative, or both:
× B) · ds over the surface of the quarter section.
(∇×
(b) B = x̂x2 − ŷy2 + ẑ2z
(c) C = r̂(sin φ )/r2 + φ̂φ(cos φ )/r2
∗ (d) D = R̂/R
r
+ ẑz
(e) E = r̂ 3 − 1+r
(f) F = (x̂y + ŷx)/(x2 + y2 )
(g) G = x̂(x2 + z2 ) − ŷ(y2 + x2 ) − ẑ(y2 + z2 )
∗ (h) H = R̂(Re−R )
y
3.59 Find the Laplacian of the following scalar functions:
(a) V = 4xy2 z3
(b) V = xy + yz + zx
(0, 3)
L1
(−3, 0) L3
Figure P3.57 Problem 3.57.
∗ (c) V = 3/(x2 + y2 )
x
(d) V = 5e−r cos φ
(e) V = 10e−R sin θ
3.60 Find the Laplacian of the following scalar functions:
(a) V1 = 10r3 sin 2φ
(b) V2 = (2/R2 ) cos θ sin φ
Chapter
4
Electrostatics
Chapter Contents
4-1
4-2
4-3
4-4
4-5
TB7
4-6
TB8
4-7
4-8
TB9
4-9
4-10
4-11
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Maxwell’s Equations, 173
Charge and Current Distributions, 173
Coulomb’s Law, 176
Gauss’s Law, 180
Electric Scalar Potential, 183
Resistive Sensors, 186
Conductors, 190
Supercapacitors as Batteries, 191
Dielectrics, 197
Electric Boundary Conditions, 200
Capacitive Sensors, 203
Capacitance, 209
Electrostatic Potential Energy, 213
Image Method, 215
Chapter 4 Summary, 217
Problems, 218
1. Evaluate the electric field and electric potential due to any
distribution of electric charges.
2. Apply Gauss’s law.
3. Calculate the resistance R of any shaped object given the
electric field at every point in its volume.
4. Describe the operational principles of resistive and capacitive sensors.
5. Calculate the capacitance of two-conductor configurations.
172
4-1 Maxwell’s Equations
Electrostatics
∇ · D = ρv ,
The modern theory of electromagnetism is based on a set of
(4.2a)
× E = 0.
∇×
(4.2b)
(4.3a)
(4.1b)
∇ · B = 0,
× H = J.
∇×
(4.3b)
(4.1c)
Maxwell’s four equations separate into two uncoupled pairs
with the first pair involving only the electric field and flux E
and D and the second pair involving only the magnetic field
and flux H and B.
four fundamental relations known as Maxwell’s equations:
Magnetostatics
∇ · D = ρv ,
∂B
×E = −
∇×
,
∂t
∇ · B = 0,
∂D
×H = J+
∇×
.
∂t
(4.1a)
(4.1d)
Here E and D are the electric field intensity and flux density
interrelated by D = ε E where ε is the electrical permittivity;
H and B are magnetic field intensity and flux density interrelated by B = µ H where µ is the magnetic permeability;
ρv is the electric charge density per unit volume; and J is
the current density per unit area. The fields and fluxes E,
D, B, H were introduced in Section 1-3, and ρv and J will
be discussed in Section 4-2. Maxwell’s equations hold in
any material, including free space (vacuum). In general, all
of the above quantities may depend on spatial location and
time t. In the interest of readability, we will not, however,
explicitly reference these dependencies (as in E(x, y, z,t))
except when the context calls for it. By formulating these
equations, published in a classic treatise in 1873, James Clerk
Maxwell established the first unified theory of electricity and
magnetism. His equations are deduced from experimental
observations reported by Coulomb, Gauss, Ampère, Faraday,
and others; they not only encapsulate the connection between
the electric field and electric charge and between the magnetic
field and electric current but also capture the bilateral coupling
between electric and magnetic fields and fluxes. Together with
some auxiliary relations, Maxwell’s equations comprise the
fundamental tenets of electromagnetic theory.
◮ Electric and magnetic fields become decoupled under
static conditions. ◭
This allows us to study electricity and magnetism as two
distinct and separate phenomena as long as the spatial distributions of charge and current flow remain constant in time.
We refer to the study of electric and magnetic phenomena
under static conditions as electrostatics and magnetostatics,
respectively. Electrostatics is the subject of the present chapter,
and we learn about magnetostatics in Chapter 5. The experience gained through studying electrostatic and magnetostatic
phenomena will prove invaluable in tackling the more involved
material in subsequent chapters that deal with time-varying
fields, charge densities, and currents.
We study electrostatics not only as a prelude to the study
of time-varying fields but also because it is an important field
in its own right. Many electronic devices and systems are
based on the principles of electrostatics. They include x-ray
machines, oscilloscopes, ink-jet electrostatic printers, liquid crystal displays, copy machines, micro-electromechanical
switches and accelerometers, and many solid-state–based control devices. Electrostatic principles also guide the design of
medical diagnostic sensors, such as the electrocardiogram,
which records the heart’s pumping pattern, and the electroencephalogram, which records brain activity, as well as the
development of numerous industrial applications.
◮ Under static conditions, none of the quantities appearing in Maxwell’s equations are functions of time (i.e.,
∂ /∂ t = 0). This happens when all charges are permanently fixed in space. If they move, they do so at a steady
rate so that ρv and J are constant in time. ◭
4-2 Charge and Current Distributions
In electromagnetics, we encounter various forms of electric
charge distributions. When put in motion, these charge distributions constitute current distributions. Charges and currents
may be distributed over a volume of space, across a surface, or
along a line.
Under these circumstances, the time derivatives of B and D in
Eqs. (4.1b) and (4.1d) vanish, and Maxwell’s equations reduce
to the following pairs.
173
174
CHAPTER 4
4-2.1 Charge Densities
z
At the atomic scale, the charge distribution in a material
is discrete, meaning that charge exists only where electrons
and nuclei are and nowhere else. In electromagnetics, we
usually are interested in studying phenomena at a much larger
scale, typically three or more orders of magnitude greater than
the spacing between adjacent atoms. At such a macroscopic
scale, we can disregard the discontinuous nature of the charge
distribution and treat the net charge contained in an elemental
volume ∆υ as if it were uniformly distributed within. Accordingly, we define the volume charge density ρv as
∆q
dq
=
∆υ →0 ∆υ
dυ
ρv = lim
(C/m3 ),
Q=
υ
ρv d υ
(C).
(4.5)
In some cases, particularly when dealing with conductors,
electric charge may be distributed across the surface of a
material, where the quantity of interest is the surface charge
density ρs , which is defined as
∆q dq
=
ρs = lim
∆s→0 ∆s
ds
(C/m2 ),
(4.6)
Line Charge Distribution
Calculate the total charge Q contained in a cylindrical tube
oriented along the z axis, as shown in Fig. 4-1(a). The line
charge density is ρℓ = 2z, where z is the distance in meters
from the bottom end of the tube. The tube length is 10 cm.
Line charge ρl
y
x
(a) Line charge distribution
z
Surface charge ρs
3 cm
ϕ
y
r
x
(b) Surface charge distribution
Figure 4-1 Charge distributions for Examples 4-1 and 4-2.
Solution: The total charge Q is
Q=
where ∆q is the charge present across an elemental surface
area ∆s. Similarly, if the charge is, for all practical purposes,
confined to a line, which need not be straight, we characterize
its distribution in terms of the line charge density ρℓ , defined
as
∆q dq
ρℓ = lim
=
(C/m).
(4.7)
∆l→0 ∆l
dl
Example 4-1:
10 cm
(4.4)
where ∆q is the charge contained in ∆υ . In general, ρv depends on spatial location (x, y, z) and t; thus, ρv = ρv (x, y, z,t).
Physically, ρv represents the average charge per unit volume
for a volume ∆υ centered at (x, y, z) with ∆υ being large
enough to contain a large number of atoms, yet it is small
enough to be regarded as a point at the macroscopic scale
under consideration. The variation of ρv with spatial location is
called its spatial distribution (or simply its distribution). The
total charge contained in volume υ is
Z
ELECTROSTATICS
Z 0.1
0
ρℓ dz =
Z 0.1
2z dz = z2
0
0.1
0
= 10−2 C.
Example 4-2: Surface Charge Distribution
The circular disk of electric charge shown in Fig. 4-1(b)
is characterized by an azimuthally symmetric surface charge
density that increases linearly with r from zero at the center to
6 C/m2 at r = 3 cm. Find the total charge present on the disk
surface.
Solution: Since ρs is symmetrical with respect to the azimuth
angle φ , it depends only on r and is given by
ρs =
6r
= 2 × 102r
3 × 10−2
(C/m2 ),
where r is in meters. In polar coordinates, an elemental area
is ds = r dr d φ , and for the disk shown in Fig. 4-1(b), the
4-2 CHARGE AND CURRENT DISTRIBUTIONS
175
limits of integration are from 0 to 2π (rad) for φ and from 0 to
3 × 10−2 m for r. Hence,
Q=
Z
S
ρs ds =
Z 2π Z 3×10−2
φ =0 r=0
(2 × 102r)r dr d φ
r3
= 2π × 2 × 10
3
u
∆q' = ρvu ∆s' ∆t
3×10−2
2
∆l
= 11.31 (mC).
(a)
0
Exercise 4-1: A square plate residing in the x–y plane
EM
∆s
ρv
is situated in the space defined by −3 m ≤ x ≤ 3 m and
−3 m ≤ y ≤ 3 m. Find the total charge on the plate if the
surface charge density is ρs = 4y2 (µ C/m2 ).
Answer: Q = 0.432 (mC). (See
∆s'
Volume charge ρv
∆s = nˆ ∆s
θ
u
.)
∆q = ρvu • ∆s ∆t
= ρvu ∆s ∆t cos θ
(b)
Exercise 4-2: A thick spherical shell centered at the origin
extends between R = 2 cm and R = 3 cm. If the volume
charge density is ρv = 3R × 10−4 (C/m3 ), find the total
charge contained in the shell.
Answer: Q = 0.61 (nC). (See
EM
.)
Figure 4-2 Charges with velocity u moving through a cross
section ∆s′ in (a) and ∆s in (b).
through it is
4-2.2 Current Density
I=
Z
S
Consider a tube with volume charge density ρv (Fig. 4-2(a)).
The charges in the tube move with velocity u along the tube
axis. Over a period ∆t, the charges move a distance ∆l = u ∆t.
The amount of charge that crosses the tube’s cross-sectional
surface ∆s′ in time ∆t is therefore
∆q′ = ρv ∆υ = ρv ∆l ∆s′ = ρv u ∆s′ ∆t.
(4.8)
Now consider the more general case where the charges are
flowing through a surface ∆s with normal n̂ not necessarily
parallel to u (Fig. 4-2(b)). In this case, the amount of charge ∆q
flowing through ∆s is
∆q = ρv u · ∆s ∆t,
(4.9)
where ∆s = n̂ ∆s and the corresponding total current flowing
in the tube is
∆I =
∆q
= ρv u · ∆s = J · ∆s.
∆t
(4.10)
J · ds
(A).
(4.12)
◮ When a current is due to the actual movement of electrically charged matter, it is called a convection current,
and J is called a convection current density. ◭
A wind-driven charged cloud, for example, gives rise to a convection current. In some cases, the charged matter constituting
the convection current consists solely of charged particles,
such as the electron beam of a scanning electron microscope
or the ion beam of a plasma propulsion system.
When a current is due to the movement of charged particles
relative to their host material, J is called a conduction current
density. In a metal wire, for example, there are equal amounts
of positive charges (in atomic nuclei) and negative charges (in
the electron shells of the atoms). None of the positive charges
and few of the negative charges can move; only those electrons
in the outermost electron shells of the atoms can be pushed
from one atom to the next if a voltage is applied across the
ends of the wire.
Here
J = ρv u
(A/m2 )
(4.11)
is defined as the current density in ampere per square meter.
Generalizing to an arbitrary surface S, the total current flowing
◮ This movement of electrons from atom to atom constitutes a conduction current. The electrons that emerge
from the wire are not necessarily the same electrons that
entered the wire at the other end. ◭
176
CHAPTER 4
ELECTROSTATICS
Conduction current, which is discussed in more detail in
Section 4-6, obeys Ohm’s law, whereas convection current
does not.
P
ˆ
R
What happens to Maxwell’s
equations under static conditions?
Concept Question 4-1:
E
R
+q
How is the current density J
related to the volume charge density ρv ?
Concept Question 4-2:
What is the difference between
convection and conduction currents?
Concept Question 4-3:
Figure 4-3 Electric field lines due to a charge q.
4-3 Coulomb’s Law
with
One of the primary goals of this chapter is to develop dexterity
in applying the expressions for the electric field intensity E
and associated electric flux density D induced by a specified
distribution of charge. Our discussion will be limited to electrostatic fields induced by stationary charge densities.
We begin by reviewing the expression for the electric
field introduced in Section 1-3.2 on the basis of the results
of Coulomb’s experiments on the electrical force between
charged bodies. Coulomb’s law, which was first introduced for
electrical charges in air and later generalized to material media,
implies that:
(1) An isolated charge q induces an electric field E at every
point in space, and at any specific point P, E is given by
E = R̂
q
4πε R2
(V/m),
(4.13)
where R̂ is a unit vector pointing from q to P (Fig. 4-3),
R is the distance between them, and ε is the electrical
permittivity of the medium containing the observation
point P.
(2) In the presence of an electric field E at a given point in
space that may be due to a single charge or a distribution
of charges, the force acting on a test charge q′ when
placed at P is
F = q′ E
(N).
(4.14)
With F measured in newtons (N) and q′ in coulombs (C), the
unit of E is (N/C), which will be shown later in Section 4-5 to
be the same as volt per meter (V/m).
For a material with electrical permittivity ε , the electric field
quantities D and E are related by
D = εE
(4.15)
ε = εr ε0 ,
(4.16)
where
ε0 = 8.85 × 10−12 ≈ (1/36π ) × 10−9
(F/m)
is the electrical permittivity of free space and εr = ε /ε0 is
called the relative permittivity (or dielectric constant) of the
material. For most materials and under a wide range of conditions, ε is independent of both the magnitude and direction
of E [as implied by Eq. (4.15)].
◮ If ε is independent of the magnitude of E, then the
material is said to be linear because D and E are related
linearly, and if it is independent of the direction of E, the
material is said to be isotropic. ◭
Materials usually do not exhibit nonlinear permittivity behavior except when the amplitude of E is very high (at levels
approaching dielectric breakdown conditions discussed later in
Section 4-7), and anisotropy is present only in certain materials
with peculiar crystalline structures. Hence, except for unique
materials under very special circumstances, the quantities D
and E are effectively redundant; for a material with known ε ,
knowledge of either D or E is sufficient to specify the other in
that material.
4-3.1
Electric Field Due to Multiple Point
Charges
The expression given by Eq. (4.13) for the field E due to
a single point charge can be extended to multiple charges.
We begin by considering two point charges, q1 and q2 , with
position vectors R1 and R2 (measured from the origin in
4-3 COULOMB’S LAW
177
z
Example 4-3:
E
E2
q1
R1
R–R
E1
1
Two point charges with
P
q1 = 2 × 10−5 C
R – R2
and
R
q2 = −4 × 10−5 C
q2
R2
y
x
Figure 4-4 The electric field E at P due to two charges is equal
to the vector sum of E1 and E2 .
Fig. 4-4). The electric field E is to be evaluated at a point P
with position vector R. At P, the electric field E1 due to q1
alone is given by Eq. (4.13) with R, which is the distance
between q1 and P, replaced with |R − R1 | and the unit vector R̂
replaced with (R − R1 )/|R − R1 |. Thus,
q1 (R − R1 )
E1 =
4πε |R − R1 |3
(V/m).
q2 (R − R2 )
4πε |R − R2 |3
(V/m).
(4.17a)
(4.17b)
◮ The electric field obeys the principle of linear superposition. ◭
Hence, the total electric field E at P due to q1 and q2 together
is determined as
E = E1 + E2
1 q1 (R − R1 ) q2 (R − R2 )
.
+
=
4πε |R − R1 |3
|R − R2 |3
(4.18)
Generalizing the preceding result to the case of N point
charges, the electric field E at point P with position vector R
due to charges q1 , q2 , . . . , qN located at points with position
vectors R1 , R2 , . . . , RN equals the vector sum of the electric
fields induced by all the individual charges. Thus,
1 N qi (R − Ri )
E=
∑ |R − Ri |3
4πε i=1
are located in free space at points with Cartesian coordinates
(1, 3, −1) and (−3, 1, −2), respectively. Find (a) the electric
field E at (3, 1, −2) and (b) the force on a 8 × 10−5 C charge
located at that point. All distances are in meters.
Solution: (a) From Eq. (4.18), the electric field E with ε = ε0
(free space) is
(R − R1 )
1
(R − R2 )
q1
E=
(V/m).
+ q2
4πε0
|R − R1 |3
|R − R2 |3
The vectors R1 , R2 , and R are
R1 = x̂ + ŷ3 − ẑ,
R2 = −x̂3 + ŷ − ẑ2,
R = x̂3 + ŷ − ẑ2.
Hence,
1
2(x̂2 − ŷ2 − ẑ) 4(x̂6)
E=
× 10−5
−
4πε0
27
216
Similarly, the electric field at P due to q2 alone is
E2 =
Electric Field Due to Two
Point Charges
=
x̂ − ŷ4 − ẑ2
× 10−5
108πε0
(V/m).
(b) The force on q3 is
x̂ − ŷ4 − ẑ2
× 10−5
108πε0
x̂2 − ŷ8 − ẑ4
× 10−10
(N).
=
27πε0
F = q3 E = 8 × 10−5 ×
Exercise 4-3: Four charges of 10 µ C each are located in
free space at points with Cartesian coordinates (−3, 0, 0),
(3, 0, 0), (0, −3, 0), and (0, 3, 0). Find the force on a 20 µ C
charge located at (0, 0, 4). All distances are in meters.
Answer: F = ẑ0.23 N. (See
EM
.)
Exercise 4-4: Two identical charges are located on the
x axis at x = 3 and x = 7. At what point in space is the net
electric field zero?
(V/m).
(4.19)
Answer: At point (5, 0, 0). (See
EM
.)
178
CHAPTER 4
Exercise 4-5: In a hydrogen atom, the electron and proton
are separated by an average distance of 5.3 × 10−11 m.
Find the magnitude of the electrical force Fe between
the two particles, and compare it with the gravitational
force Fg between them.
If the charge is distributed across a surface S′ with surface
charge density ρs , then dq = ρs ds′ , and if it is distributed
along a line l ′ with a line charge density ρℓ , then dq = ρℓ dl ′ .
Accordingly, the electric fields due to surface and line charge
distributions are
Answer: Fe = 8.2 × 10−8 N, and Fg = 3.6 × 10−47 N.
(See
EM
ELECTROSTATICS
ρs ds′
,
R′2
S′
(surface distribution)
E=
.)
1
4πε
Z
R̂
′
(4.21b)
4-3.2 Electric Field Due to a Charge Distribution
We now extend the results obtained for the field due to discrete
point charges to continuous charge distributions. Consider a
volume υ ′ that contains a distribution of electric charge with
volume charge density ρv , which may vary spatially within υ ′
(Fig. 4-5). The differential electric field at a point P due to
a differential amount of charge dq = ρv d υ ′ contained in a
differential volume d υ ′ is
dE = R̂
′
Z
υ′
dE =
1
4πε
Z
υ′
R̂
′
1
4πε
Z
R̂
′
(4.21c)
Example 4-4: Electric Field of a Ring of Charge
′
dq
′ ρv d υ
=
R̂
,
4πε R′2
4πε R′2
(4.20)
where R′ is the vector from the differential volume d υ ′ to
point P. Applying the principle of linear superposition, the
total electric field E is obtained by integrating the fields due
to all differential charges in υ ′ . Thus,
E=
ρℓ dl ′
.
R′2
l′
(line distribution)
E=
ρv d υ ′
.
R′2
(4.21a)
(volume distribution)
A ring of charge of radius b is characterized by a uniform line
charge density of positive polarity ρℓ . The ring resides in free
space and is positioned in the x–y plane, as shown in Fig. 4-6.
Determine the electric field intensity E at a point P = (0, 0, h)
along the axis of the ring at a distance h from its center.
Solution: We start by considering the electric field generated
by a differential ring segment with cylindrical coordinates
(b, φ , 0) in Fig. 4-6(a). The segment has length dl = b d φ and
contains charge dq = ρℓ dl = ρℓ b d φ . The distance vector R′1
from segment 1 to point P = (0, 0, h) is
′
It is important to note that, in general, both R′ and R̂ vary as a
function of position over the integration volume υ ′ .
R′1 = −r̂b + ẑh,
from which it follows that
P
dE
R'
ρv dυ'
υ'
Figure 4-5 Electric field due to a volume charge distribution.
R′1 = |R′1 | =
p
b2 + h2 ,
′
R̂1 =
R′1
−r̂b + ẑh
=√
.
|R′1 |
b2 + h2
The electric field at P = (0, 0, h) due to the charge in segment 1
therefore is
dE1 =
ρℓ b (−r̂b + ẑh)
1
′ ρℓ dl
R̂
=
dφ .
4πε0 1 R′1 2
4πε0 (b2 + h2)3/2
The field dE1 has component dE1r along −r̂ and component dE1z along ẑ. From symmetry considerations, the
field dE2 generated by differential segment 2 in Fig. 4-6(b),
which is located diametrically opposite to segment 1, is identical to dE1 except that the r̂ component of dE2 is opposite that
4-3 COULOMB’S LAW
179
opposite at (φ + π ), we can obtain the total field generated by
the ring by integrating Eq. (4.22) over a semicircle as
z
dE1
dE1z
dE1r
P = (0, 0, h)
h
b
+
+
+
= ẑ
R'1
+ + +
+
E = ẑ
+
ρl
ρℓ bh
2πε0 (b2 + h2)3/2
Z π
dφ
0
h
ρℓ bh
= ẑ
Q,
2
2
3/2
2
2ε0 (b + h )
4πε0 (b + h2 )3/2
(4.23)
where Q = 2π bρℓ is the total charge on the ring.
+
+
dφ
y
φ
+ + +
1
Example 4-5:
dl = b dφ
Electric Field of a Circular
Disk of Charge
x
(a)
Find the electric field at point P with Cartesian coordinates
(0, 0, h) due to a circular disk of radius a and uniform charge
density ρs residing in the x–y plane (Fig. 4-7). Also, evaluate E
due to an infinite sheet of charge density ρs by letting a → ∞.
z
dE = dE1 + dE2
dE1
dE2
dE1r
dE2r
R'2
2
+
E
P = (0, 0, h)
R'1
+
+
+
φ+π
+
+
z
+ +
h
+
+
ρs
y
dq = 2πρsr dr
φ
+ +
1
r
x
a
y
(b)
dr
Figure 4-6 Ring of charge with line density ρℓ . (a) The
field dE1 due to infinitesimal segment 1 and (b) the fields dE1
and dE2 due to segments at diametrically opposite locations
(Example 4-4).
of dE1 . Hence, the r̂ components in the sum cancel and the ẑ
contributions add. The sum of the two contributions is
dE = dE1 + dE2 = ẑ
ρℓ bh
dφ
.
2
2πε0 (b + h2)3/2
(4.22)
Since for every ring segment in the semicircle defined over the
azimuthal range 0 ≤ φ ≤ π (the right-hand half of the circular
ring) there is a corresponding segment located diametrically
a
x
Figure 4-7 Circular disk of charge with surface charge
density ρs . The electric field at P = (0, 0, h) points along the
z direction (Example 4-5).
Solution: Building on the expression obtained in Example 4-4 for the on-axis electric field due to a circular ring of
charge, we can determine the field due to the circular disk
by treating the disk as a set of concentric rings. A ring of
radius r and width dr has an area ds = 2π r dr and contains
charge dq = ρs ds = 2πρsr dr. Upon using this expression in
180
CHAPTER 4
Eq. (4.23) and also replacing b with r, we obtain the following
expression for the field due to the ring:
dE = ẑ
ELECTROSTATICS
and propose alternative techniques for evaluating electric fields
induced by electric charge. To that end, we restate Eq. (4.1a):
h
(2πρsr dr).
4πε0 (r2 + h2 )3/2
∇ · D = ρv ,
(4.26)
(differential form of Gauss’s law)
The total field at P is obtained by integrating the expression
over the limits r = 0 to r = a:
Z
ρs h a
ρs
r dr
|h|
E = ẑ
1− √
= ±ẑ
,
2ε0 0 (r2 + h2)3/2
2ε0
a2 + h2
(4.24)
with the plus sign for h > 0 (P above the disk) and the minus
sign when h < 0 (P below the disk).
For an infinite sheet of charge with a = ∞,
E = ±ẑ
ρs
.
2ε0
(4.25)
(infinite sheet of charge)
We note that for an infinite sheet of charge E is the same at all
points above the x–y plane, and a similar statement applies for
points below the x–y plane.
When characterizing the electrical permittivity of a material, what do the terms linear
and isotropic mean?
Concept Question 4-4:
If the electric field is zero at
a given point in space, does this imply the absence of
electric charges?
Concept Question 4-5:
State the principle of linear
superposition as it applies to the electric field due to a
distribution of electric charge.
Concept Question 4-6:
Exercise 4-6: An infinite sheet with uniform surface
charge density ρs is located at z = 0 (x–y plane), and
another infinite sheet with density −ρs is located at
z = 2 m with both in free space. Determine E everywhere.
Answer: E = 0 for z < 0; E = ẑρs /ε0 for 0 < z < 2 m;
and E = 0 for z > 2 m. (See
EM
.)
4-4 Gauss’s Law
In this section, we use Maxwell’s equations to confirm the
expressions for the electric field implied by Coulomb’s law,
which is referred to as the differential form of Gauss’s law.
The adjective “differential” refers to the fact that the divergence operation involves spatial derivatives. As we see shortly,
Eq. (4.26) can be converted to an integral form. When solving
electromagnetic problems, we often go back and forth between
equations in differential and integral form, depending on which
of the two happens to be the more applicable or convenient
to use. To convert Eq. (4.26) into integral form, we multiply
both sides by d υ and evaluate their integrals over an arbitrary
volume υ :
Z
Z
∇ · D d υ = ρv d υ = Q.
(4.27)
υ
υ
Here, Q is the total charge enclosed in υ . The divergence
theorem, given by Eq. (3.98), states that the volume integral of
the divergence of any vector over a volume υ equals the total
outward flux of that vector through the surface S enclosing υ .
Thus, for the vector D,
Z
υ
∇ · D dυ =
Z
S
D · ds.
(4.28)
Comparison of Eq. (4.27) with Eq. (4.28) leads to
Z
S
D · ds = Q.
(4.29)
(integral form of Gauss’s law)
◮ The integral form of Gauss’s law is illustrated diagrammatically in Fig. 4-8; for each differential surface
element ds, D · ds is the electric field flux flowing outward
of υ through ds, and the total flux through surface S
equals the enclosed charge Q. The surface S is called a
Gaussian surface. ◭
The integral form of Gauss’s law can be applied to determine D due to a single isolated point charge q by enclosing
the latter with a closed, spherical, Gaussian surface S of
arbitrary radius R centered at q (Fig. 4-9). From symmetry
considerations and assuming that q is positive, the direction
of D must be radially outward along the unit vector R̂, and DR ,
which is the magnitude of D, must be the same at all points
on S. Thus, at any point on S,
D = R̂DR ,
(4.30)
4-4 GAUSS’S LAW
181
Total charge
in υ
to find E for any specified distribution of charge, Gauss’s law
is easier to apply than Coulomb’s law, but its utility is limited
to symmetrical charge distributions.
◮ Gauss’s law, as given by Eq. (4.29), provides a convenient method for determining the flux density D when
the charge distribution possesses symmetry properties that
allow us to infer the variations of the magnitude and
direction of D as a function of spatial location. This
facilitates the integration of D over a cleverly chosen
Gaussian surface. ◭
D • ds
Q
υ
Gaussian surface S
enclosing volume υ
Figure 4-8 The integral form of Gauss’s law states that the
outward flux of D through a surface is proportional to the
enclosed charge Q.
Because at every point on the surface the direction of ds is
along its outward normal, only the normal component of D
at the surface contributes to the integral in Eq. (4.29). To
successfully apply Gauss’s law, the surface S should be chosen
from symmetry considerations so that, across each subsurface
of S, D is constant in magnitude and its direction is either
normal or purely tangential to the subsurface. These aspects
are illustrated in Example 4-6.
R
Example 4-6:
R̂
q
Electric Field of an Infinite
Line Charge
ds
D
Gaussian surface
Figure 4-9 Electric field D due to point charge q.
and ds = R̂ ds. Applying Gauss’s law gives
Z
S
D · ds =
Z
S
R̂DR · R̂ ds =
Z
S
DR ds = DR (4π R2) = q.
(4.31)
Solving for DR and then inserting the result in Eq. (4.30) gives
the expression for the electric field E induced by an isolated
point charge in a medium with permittivity ε :
E=
q
D
= R̂
ε
4πε R2
(V/m).
(4.32)
This is identical with Eq. (4.13) obtained from Coulomb’s law;
after all, Maxwell’s equations incorporate Coulomb’s law. For
this simple case of an isolated point charge, it does not matter
whether Coulomb’s law or Gauss’s law is used to obtain the
expression for E. However, it does matter which approach we
follow when we deal with multiple point charges or continuous
charge distributions. Even though Coulomb’s law can be used
Use Gauss’s law to obtain an expression for E due to an
infinitely long line with uniform charge density ρℓ that resides
along the z axis in free space.
Solution: Since the charge density along the line is uniform,
infinite in extent, and residing along the z axis, symmetry
considerations dictate that D is in the radial r̂ direction and
cannot depend on φ or z. Thus, D = r̂ Dr . Therefore, we
construct a finite cylindrical Gaussian surface of radius r
and height h, which is concentric around the line of charge
(Fig. 4-10). The total charge contained within the cylinder is
Q = ρℓ h. Since D is along r̂, the top and bottom surfaces of
the cylinder do not contribute to the surface integral on the
left-hand side of Eq. (4.29); that is, only the curved surface
contributes to the integral. Hence,
Z h Z 2π
z=0 φ =0
or
r̂Dr · r̂r d φ dz = ρℓ h
2π hDr r = ρℓ h,
which yields
E=
D
Dr
ρℓ
= r̂
= r̂
.
ε0
ε0
2πε0 r
(infinite line charge)
(4.33)
182
CHAPTER 4
and carrying charge density ρℓ1 = 1 (nC/m), and a second one
residing in the y–z plane parallel to the y axis and carrying a
charge density ρℓ2 = −2 (nC/m). Determine the electric field
at the origin.
z
uniform line
charge ρl
r
Solution: The electric field E is the sum of two electric field
components:
h
E = E1 + E2 ,
ds
D
Gaussian surface
Figure 4-10 Gaussian surface around an infinitely long line of
charge (Example 4-6).
where E1 and E2 are the electric fields due to line charges
1 and 2, respectively. According to Eq. (4.33), the direction
of the electric field r̂ is perpendicular to the direction of the
line charge and points away from the line of charge (if ρℓ is
positive). Hence, for the first line of charge, ρℓ1 = 1 (nC/m),
r̂1 = −ŷ, r1 = 2, and
E1 =
Note that Eq. (4.33) is applicable for any infinite line of charge,
regardless of its location and direction, as long as r̂ is properly
defined as the radial distance vector from the line charge to the
observation point (i.e., r̂ is perpendicular to the line of charge).
Example 4-7:
ELECTROSTATICS
r̂1 ρℓ1
−ŷ 10−9
= −ŷ 9 V/m.
=
2πε0 r1
2π × 361π × 10−9 × 2
Similarly, for the second line of charge, ρℓ2 = −2 (nC/m),
r̂2 = −ẑ, r2 = 6, and
E2 =
Two Infinite Lines of Charge
r̂2 ρℓ2
−ẑ(−2) × 10−9
= ẑ 6
=
2πε0 r2
2π × 361π × 10−9 × 6
V/m.
Hence,
Figure 4-11 depicts the presence of two infinite lines of charge
in free space: one residing in the x–y plane parallel to the x̂ axis
E = E1 + E2 = (−ŷ 9 + ẑ 6) V/m.
Explain Gauss’s law. Under
what circumstances is it useful?
Concept Question 4-7:
z
Concept Question 4-8:
6m
ρℓ = −2 (nC/m)
rˆ 2
E
2
Exercise 4-7: Two infinite lines, each carrying a uniform
ˆ
E2 = z6
rˆ 1
ˆ
E1 = −y9
x
How should one choose a
Gaussian surface?
y
2m
ρℓ = 1 (nC/m)
1
Figure 4-11 Two infinite lines of charge (Example 4-7).
charge density ρℓ , reside in free space parallel to the z axis
at x = 1 and x = −1. Determine E at an arbitrary point
along the y axis.
Answer: E = ŷρℓ y/ πε0 (y2 + 1) . (See EM .)
Exercise 4-8: A thin spherical shell of radius a carries
a uniform surface charge density ρs . Use Gauss’s law to
determine E everywhere in free space.
(
0
for R < a;
(See EM .)
Answer: E =
R̂ρs a2 /(ε R2 ) for R > a.
4-5 ELECTRIC SCALAR POTENTIAL
183
Exercise 4-9: A spherical volume of radius a contains a
uniform volume charge density ρv . Use Gauss’s law to
determine D for (a) R ≤ a and (b) R ≥ a.
Answer: (a) D = R̂ρv R/3,
(b) D = R̂ρv a3 /(3R2 ). (See
EM
an external force Fext to counteract Fe , which requires the
expenditure of energy. To move q without acceleration (at
constant speed), the net force acting on the charge must be
zero, which means that Fext + Fe = 0, or
Fext = −Fe = −qE.
.)
(4.34)
The work done (or energy expended) in moving any object a
vector differential distance dl while exerting a force Fext is
dW = Fext · dl = −qE · dl
4-5 Electric Scalar Potential
The operation of an electric circuit usually is described in
terms of the currents flowing through its branches and the
voltages at its nodes. The voltage difference V between two
points in a circuit represents the amount of work, or potential
energy, required to move a unit charge from one to the other.
4-5.1 Electric Potential as a Function of Electric
Field
dW = −q(−ŷE) · ŷ dy = qE dy.
(4.36)
The differential electric potential energy dW per unit charge
is called the differential electric potential (or differential
voltage) dV . That is,
dV =
dW
= −E · dl
q
(J/C or V).
(4.37)
The unit of V is the volt (V) with 1 V = 1 J/C, and since V
is measured in volts, the electric field is expressed in volts per
meter (V/m).
The potential difference corresponding to moving a point
charge from point P1 to point P2 (Fig. 4-13) is obtained by
integrating Eq. (4.37) along any path between them. That is,
Z P2
dV = −
Z P2
E · dl,
(4.38)
P1
V21 = V2 − V1 = −
Z P2
E · dl,
(4.39)
P1
or
We begin by considering the simple case of a positive charge q
in a uniform electric field E = −ŷE in the −y direction
(Fig. 4-12). The presence of the field E exerts a force Fe = qE
on the charge in the −y direction. To move the charge along the
positive y direction (against the force Fe ), we need to provide
P1
E
y
E
C1
P2
path 1
Fext
C2
dy
path 2
q
E
(4.35)
Work (or energy) is measured in joules (J). If the charge is
moved a distance dy along ŷ, then
◮ The term “voltage” is short for “voltage potential” and
synonymous with electric potential. ◭
Even though when analyzing a circuit we may not consider the
electric fields present in the circuit, it is in fact the existence of
these fields that gives rise to voltage differences across circuit
elements such as resistors or capacitors. The relationship
between the electric field E and the electric potential V is the
subject of this section.
(J).
E
E
Fe
E
P1
path 3
C3
x
Figure 4-13 In electrostatics, the potential difference between
Figure 4-12 Work done in moving a charge q a distance dy
against the electric field E is dW = qE dy.
P2 and P1 is the same irrespective of the path used for calculating
the line integral of the electric field between them.
184
CHAPTER 4
where V1 and V2 are the electric potentials at points P1 and P2 ,
respectively. The result of the line integral on the right-hand
side of Eq. (4.39) is independent of the specific integration
path that connects points P1 and P2 . This follows immediately
from the law of conservation of energy. To illustrate with an
example, consider a particle in Earth’s gravitational field. If
the particle is raised from a height h1 above Earth’s surface
to height h2 , the particle gains potential energy in an amount
proportional to (h2 − h1 ). If instead we were to first raise the
particle from height h1 to a height h3 greater than h2 , thereby
giving it potential energy proportional to (h3 − h1 ), and then
let it drop back to height h2 by expending an energy amount
proportional to (h3 − h2), its net gain in potential energy would
again be proportional to (h2 − h1).
The same principle applies to the electric potential energy W
and to the potential difference (V2 −V1 ). The voltage difference
between two nodes in an electric circuit has the same value
regardless of which path in the circuit we follow between the
nodes. Moreover, Kirchhoff’s voltage law states that the net
voltage drop around a closed loop is zero. If we go from
P1 to P2 by path 1 in Fig. 4-13 and then return from P2 to
P1 by path 2, the right-hand side of Eq. (4.39) becomes a
closed contour, and the left-hand side vanishes. In fact, the line
integral of the electrostatic field E around any closed contour
C is zero:
Z
C
E · dl = 0.
(electrostatics)
(4.40)
◮ A vector field whose line integral along any closed path
is zero is called a conservative or an irrotational field.
Hence, the electrostatic field E is conservative. ◭
As we will see later in Chapter 6, if E is a time-varying
function, it is no longer conservative, and its line integral along
a closed path is not necessarily zero.
The conservative property of the electrostatic field can
be deduced from Maxwell’s second equation, Eq. (4.1b). If
∂ /∂ t = 0, then
× E = 0.
∇×
(4.41)
× E over an open surface S
If we take the surface integral of ∇×
and then apply Stokes’s theorem expressed by Eq. (3.107) to
convert the surface integral into a line integral, we obtain
Z
S
× E) · ds =
(∇×
Z
C
E · dl = 0,
(4.42)
where C is a closed contour surrounding S. Thus, Eq. (4.41) is
the differential-form equivalent of Eq. (4.40).
ELECTROSTATICS
We now define what we mean by the electric potential V at
a point in space. Before we do so, however, let us revisit our
electric-circuit analogue. Just as a node in a circuit cannot be
assigned an absolute voltage, a point in space cannot have an
absolute electric potential. The voltage of a node in a circuit is
measured relative to that of a conveniently chosen reference
point to which we have assigned a voltage of zero, which
we call ground. The same principle applies to the electric
potential V . Usually (but not always), the reference point is
chosen to be at infinity. That is, in Eq. (4.39) we assume that
V1 = 0 when P1 is at infinity. Therefore, the electric potential V
at any point P is
V =−
Example 4-8:
Z P
∞
E · dl
(V).
(4.43)
Computing V from E along
Two Paths
A vector field is said to be conservative if its line integral
between two points is the same—irrespective of the path taken
between them. In a given region of space, the field E is given
by
E = x̂ x2 + ŷ y2 + ẑ z2 .
(4.44)
(a) Confirm that E is conservative by demonstrating that
∇ × E = 0. (b) Compute the potential difference V21 between
points 1 and 2 in Fig. 4-14 following the direct path
between them. (c) Compute V21 by following the path ABCD
between points 1 and 2.
Solution: (a) The given electric field has components
Ex = x2 , Ey = y2 , and Ez = z2 . Applying the curl operator to
E gives
x̂
∂
∇×E =
∂x
x2
2
∂z
= x̂
∂y
ŷ
∂
∂y
y2
−
ẑ
∂
∂z
z2
2
∂y
∂z
− ŷ
∂ z2 ∂ x2
−
∂x
∂z
= x̂(0 − 0) − ŷ(0 − 0) + ẑ(0 − 0) = 0.
(b) Voltage V21 is given by
V21 = −
Z P2
P1
E · dl.
+ ẑ
∂ y2 ∂ x2
−
∂x
∂y
4-5 ELECTRIC SCALAR POTENTIAL
185
The potential difference is then
y
P2(3,2,0)
D
2
=−
Path 12
1
1
2
V21 = −
Path ABCD
3
B
C
=−
=−
x
Z P2
E · dl
P1
Z x=3
x=1
Z 3
x=1
Z 3
x=1
[x̂ x2 + ŷ(2x − 4)2] · [x̂ dx + ŷ 2 dx]
[x2 + 2(2x − 4)2] dx
(9x2 − 32x + 32) dx = −14
(V).
(4.49)
(c) Path ABCD in Fig. 4-14 consists of three segments.
A to B:
−1
E = x̂ x2 + ŷ y2 + ẑ z2
−2
A
x=1, z=0
= x̂ 1 + ŷ y2 ,
(4.50a)
and
P1(1,−2,0)
dl = ŷ dy.
(4.50b)
B to C:
Figure 4-14 Computing V21 along two paths (Example 4-8).
E = x̂ x2 + ŷ y2 + ẑ z2
y=0, z=0
= x̂ x2 ,
(4.51a)
and
The straight-line path resides in the x–y plane, so it is described
by the linear form y = ax + b. At point 1, x1 = 1 and y1 = −2.
Hence,
−2 = a + b.
dl = x̂ dx.
C to D:
E = x̂ x2 + ŷ y2 + ẑ z2
= x̂ 9 + ŷ y2 ,
(4.52a)
dl = ŷ dy.
(4.52b)
Hence,
2 = 3a + b.
V21 = −
The two equations lead to a = 2, b = −4, and
y = 2x − 4.
(4.45)
Since path P1 –P2 is entirely in the x–y plane, we can set z = 0
in the expression for E. Also, we can use the relation given by
Eq. (4.45) to reduce E to a single variable:
z=0 and y=2x−4
x=3, z=0
and
Similarly, at point 2, x2 = 3 and y2 = 2, so
E = x̂ x2 + ŷ y2 + ẑ z2
(4.51b)
= x̂ x2 + ŷ(2x − 4)2.
(4.46)
In general,
P1
"Z
E · dl
B@x=1, y=0
A@x=1, y=−2
+
(x̂ 1 + ŷ y2 ) · ŷ dy
Z C@x=3, y=0
B@x=1, y=0
+
Z D@x=3, y=2
C@x=3, y=0
x̂ x2 · x̂ dx
(x̂ 9 + ŷ y2 ) · ŷ dy
#
0
3
2
x3
y3
y3
=−
+
+
3 −2
3 1
3 0
8 27 1 8
= −14
− +
=− + +
3
3
3 3
#
"
dl = x̂ dx + ŷ dy + ẑ dz.
In the x–y plane, dz = 0, and along the straight-line path given
by y = 2x − 4,
dy = 2 dx.
(4.47)
Hence,
dl = x̂ dx + ŷ 2 dx.
=−
Z P2
(4.48)
(V),
(4.53)
which is identical with the result given by Eq. (4.49) for
the line integral along the straight-line path between points 1
and 2.
186
CHAPTER 4
ELECTROSTATICS
Technology Brief 7: Resistive Sensors
About 30 electric/electronic systems and
more than 100
sensors
An electrical sensor is a device capable of responding
to an applied stimulus by generating an electrical signal
whose voltage, current, or some other attribute is related
to the intensity of the stimulus.
◮ The family of possible stimuli encompasses a wide
array of physical, chemical, and biological quantities,
including temperature, pressure, position, distance,
motion, velocity, acceleration, concentration (of a
gas or liquid), blood flow, etc. ◭
ABS TPM ABC ZV ESP ECT LWR
PTS AAC RCU DTR CDI
System
Distronic
Electronic controlled transmission
Roof control unit
Antilock braking system
Central locking system
Dyn. beam levelling
Common-rail diesel injection
Automatic air condition
Active body control
Tire pressure monitoring
Electron. stability program
Parktronic system
The sensing process relies on measuring resistance,
capacitance, inductance, induced electromotive force
(emf), oscillation frequency or time delay, among others. Sensors are integral to the operation of just about
every instrument that uses electronic systems, from
automobiles and airplanes to computers and cell phones
(Fig. TF7-1). This Technology Brief covers resistive
sensors. Capacitive, inductive, and emf sensors are
covered separately (here and in later chapters).
Abbrev.
DTR
ECT
RCU
ABS
ZV
LWR
CDI
AAC
ABC
TPM
ESP
PTS
Sensors
3
9
7
4
3
6
11
13
12
11
14
12
Figure TF7-1 Most cars use on the order of 100 sensors.
Piezoresistivity
According to Eq. (4.70), the resistance of a cylindrical
resistor or wire conductor is given by R = l/σ A, where l
is the cylinder’s length, A is its cross-sectional area, and
σ is the conductivity of its material. Stretching the wire
by an applied external force causes l to increase and
A to decrease. Consequently, R increases (Fig. TF7-2).
Conversely, compressing the wire causes R to decrease.
The Greek word piezein means to press, from which
the term piezoresistivity is derived. This should not be
confused with piezoelectricity, which is an emf effect.
(See EMF Sensors in Technology Brief 12.)
The relationship between the resistance R of a
piezoresistor and the applied force F can be modeled
by the approximate linear equation
αF
R = R0 1 +
,
A0
where R0 is the unstressed resistance (@ F = 0), A0 is the
unstressed cross-sectional area of the resistor, and α is
the piezoresistive coefficient of the resistor material.
The force F is positive if it is causing the resistor to
stretch and negative if it is compressing it.
R (Ω)
Stretching
F
F
F
F
Compression
Force (N)
F=0
Figure TF7-2 Piezoresistance varies with applied force.
4-5 ELECTRIC SCALAR POTENTIAL
Film
z
Flat
F=0
Stretched
Figure TF7-3 Piezoresistor films.
An elastic resistive sensor is well suited for measuring
the deformation z of a surface (Fig. TF7-3), which can be
related to the pressure applied to the surface; and if z is
recorded as a function of time, it is possible to derive
the velocity and acceleration of the surface’s motion.
To realize high longitudinal piezoresistive sensitivity (the
ratio of the normalized change in resistance, ∆R/R0 , to
the corresponding change in length, ∆l/l0 , caused by
the applied force), the piezoresistor is often designed
as a serpentine-shaped wire (Fig. TF7-4(a)) bonded on
a flexible plastic substrate and glued onto the surface
whose deformation is to be monitored. Copper and nickel
alloys are commonly used for making the sensor wires,
although in some applications silicon is used instead
(Fig. TF7-4(b)) because it has a very high piezoresistive
sensitivity.
187
◮ By connecting the piezoresistor to a Wheatstone
bridge circuit (Fig. TF7-5) where the other three
resistors are all identical in value and equal to R0
(the resistance of the piezoresistor when no external
force is present), the voltage output becomes directly
proportional to the normalized resistance change:
∆R/R0 . ◭
V0
R0
V0
+
−
R0
V1
Vout
R0
V0
Vout =
4
V2
R0 + ∆R
Flexible
resistor
∆R
R0
Figure TF7-5 Wheatstone bridge circuit with piezoresistor.
Ohmic
contacts
Metal wire
(a) Serpentine wire
Silicon
piezoresistor
(b) Silicon piezoresistor
Figure TF7-4 Metal and silicon piezoresistors.
188
CHAPTER 4
4-5.2 Electric Potential Due to Point Charges
The electric field due to a point charge q located at the origin
is given by Eq. (4.32) as
q
E = R̂
4πε R2
(V/m).
(4.54)
The field is radially directed and decays quadratically with the
distance R from the observer to the charge.
As was stated earlier, the choice of integration path between
the end points in Eq. (4.43) is arbitrary. Hence, we can
conveniently choose the path to be along the radial direction R̂,
in which case dl = R̂ dR and
Z R
q
q · R̂ dR =
(V).
(4.55)
V =−
R̂
2
4πε R
4πε R
∞
If the charge q is at a location other than the origin, say at
position vector R1 , then V at observation position vector R
becomes
q
V=
(V),
(4.56)
4πε |R − R1 |
where |R − R1 | is the distance between the observation point
and the location of the charge q. The principle of superposition
applied previously to the electric field E also applies to
the electric potential V . Hence, for N discrete point charges
q1 , q2 , . . . , qN residing at position vectors R1 , R2 , . . . , RN , the
electric potential is
V=
1
4πε
N
qi
∑ |R − Ri |
(V).
(4.57)
i=1
To obtain expressions for the electric potential V due to
continuous charge distributions over a volume υ ′ , across a
surface S ′ , or along a line l ′ , we (1) replace qi in Eq. (4.57)
with ρv d υ ′ , ρs ds′ , and ρℓ dl ′ , respectively; (2) convert the
summation into an integration; and (3) define R′ = |R − Ri | as
the distance between the integration point and the observation
point. These steps lead to the following expressions:
1
ρv
d υ ′ (volume distribution),
4πε υ ′ R′
Z
1
ρs ′
ds (surface distribution),
V=
4πε S′ R′
Z
ρℓ ′
1
dl (line distribution).
V=
4πε l ′ R′
Z
Electric Field as a Function of Electric
Potential
In Section 4-5.1, we expressed V in terms of a line integral over E. Now we explore the inverse relationship by
reexamining Eq. (4.37):
dV = −E · dl.
(4.59)
dV = ∇V · dl,
(4.60)
E = −∇V .
(4.61)
For a scalar function V , Eq. (3.73) gives
where ∇V is the gradient of V . Comparison of Eq. (4.59) with
Eq. (4.60) leads to
◮ This differential relationship between V and E allows
us to determine E for any charge distribution by first
calculating V and then taking the negative gradient of V
to find E. ◭
The expressions for V , given by Eqs. (4.57) to (4.58c), involve
scalar sums and scalar integrals, and as such are usually
much easier to evaluate than the vector sums and integrals
in the expressions for E derived in Section 4-3 on the basis
of Coulomb’s law. Thus, even though the electric potential
approach for finding E is a two-step process, it is conceptually
and computationally simpler to apply than the direct method
based on Coulomb’s law.
Example 4-9: Electric Field of an Electric Dipole
4-5.3 Electric Potential Due to Continuous
Distributions
V=
4-5.4
ELECTROSTATICS
(4.58a)
(4.58b)
(4.58c)
An electric dipole consists of two point charges of equal
magnitude but opposite polarity separated by a distance d
(Fig. 4-15(a)). Determine V and E at any point P given that
P is at a distance R ≫ d from the dipole center and the dipole
resides in free space.
Solution: To simplify the derivation, we align the dipole
along the z axis and center it at the origin (Fig. 4-15(a)). For the
two charges shown in Fig. 4-15(a), application of Eq. (4.57)
gives
1
q
R2 − R1
−q
q
V=
+
=
.
4πε0 R1 R2
4πε0
R1 R2
Since d ≪ R, the lines labeled R1 and R2 in Fig. 4-15(a)
are approximately parallel to each other, in which case the
following approximations apply:
R2 − R1 ≈ d cos θ ,
R1 R2 ≈ R2 .
4-5 ELECTRIC SCALAR POTENTIAL
189
P = (R, θ, φ)
z
where we have used the expression for ∇V in spherical
coordinates given in Appendix C. Upon taking the derivatives
of the expression for V given by Eq. (4.62) with respect to R
and θ and then substituting the results in Eq. (4.65), we obtain
R1
R
+q
E=
R2
θ
d
–q
y
qd
(R̂ 2 cos θ + θ̂θ sin θ ) (V/m).
4πε0 R3
(4.66)
We stress that the expressions for V and E given by Eqs. (4.64)
and (4.66) apply only when R ≫ d. To compute V and E
at points in the vicinity of the two dipole charges, it is
necessary to perform all calculations without resorting to the
far-distance approximations that led to Eq. (4.62). Such an
exact calculation for E leads to the field pattern shown in
Fig. 4-15(b).
d cos θ
x
(a) Electric dipole
4-5.5 Poisson’s Equation
E
With D = ε E, the differential form of Gauss’s law given by
Eq. (4.26) may be cast as
∇·E =
(4.67)
Inserting Eq. (4.61) in Eq. (4.67) gives
(b) Electric-field pattern
Figure 4-15
ρv
.
ε
∇ · (∇V ) = −
Electric dipole with dipole moment p = qd
(Example 4-9).
ρv
.
ε
(4.68)
Given Eq. (3.110) for the Laplacian of a scalar function V ,
Hence,
qd cos θ
V=
.
(4.62)
4πε0 R2
To generalize this result to an arbitrarily oriented dipole, note
that the numerator of Eq. (4.62) can be expressed as the dot
product of qd (where d is the distance vector from −q to +q)
and the unit vector R̂ pointing from the center of the dipole
toward the observation point P. That is,
qd cos θ = qd · R̂ = p · R̂,
(4.63)
where p = qd is called the dipole moment. Using Eq. (4.63) in
Eq. (4.62) then gives
p · R̂
V=
4πε0 R2
∇2V = ∇ · (∇V ) =
∇2V = −
ρv
ε
(Poisson’s equation).
(4.70)
This is known as Poisson’s equation. For a volume υ ′ containing a volume charge density distribution ρv , the solution
for V derived previously and expressed by Eq. (4.58a) as
(4.64)
In spherical coordinates, Eq. (4.61) is given by
∂V
∂V
1 ∂V
1
, (4.65)
E = −∇V = − R̂
+ φ̂φ
+ θ̂θ
∂R
R ∂θ
R sin θ ∂ φ
(4.69)
Eq. (4.68) can be cast in the abbreviated form
V=
(electric dipole).
∂ 2V ∂ 2V ∂ 2V
+ 2 + 2 ,
∂ x2
∂y
∂z
1
4πε
Z
υ′
ρv
dυ ′
R′
(4.71)
satisfies Eq. (4.70). If the medium under consideration contains no charges, Eq. (4.70) reduces to
∇2V = 0
(Laplace’s equation),
(4.72)
190
CHAPTER 4
and it is then referred to as Laplace’s equation. Poisson’s and
Laplace’s equations are useful for determining the electrostatic
potential V in regions with boundaries on which V is known,
such as the region between the plates of a capacitor with a
specified voltage difference across it.
Concept Question 4-9:
What is a conservative field?
Why is the electric potential
at a point in space always defined relative to the potential
at some reference point?
Concept Question 4-10:
Explain why Eq. (4.40) is a
mathematical statement of Kirchhoff’s voltage law.
Concept Question 4-11:
Why is it usually easier to
compute V for a given charge distribution and then find
E using E = −∇V than to compute E directly by applying
Coulomb’s law?
ELECTROSTATICS
◮ The conductivity of a material is a measure of how
easily electrons can travel through the material under the
influence of an externally applied electric field. ◭
Materials are classified as conductors (metals) or dielectrics
(insulators) according to the magnitudes of their conductivities. A conductor has a large number of loosely attached
electrons in the outermost shells of its atoms. In the absence
of an external electric field, these free electrons move in
random directions and with varying speeds. Their random
motion produces zero average current through the conductor.
Upon applying an external electric field, however, the electrons
migrate from one atom to the next in the direction opposite
that of the external field. Their movement gives rise to a
conduction current density
Concept Question 4-12:
Concept Question 4-13:
What is an electric dipole?
Exercise 4-10: Determine the electric potential at the
origin due to four 20 µ C charges residing in free space
at the corners of a 2 m × 2 m square centered about the
origin in the x–y plane.
√
Answer: V = 2 × 10−5 /(πε0 ) (V). (See EM .)
Exercise 4-11: A spherical shell of radius a has a uniform
surface charge density ρs . Determine (a) the electric
potential and (b) the electric field with both at the center
of the shell.
Answer: (a) V = ρs a/ε (V), (b) E = 0. (See
EM
.)
4-6 Conductors
The electromagnetic constitutive parameters of a material
medium are its electrical permittivity ε , magnetic permeability µ , and conductivity σ . A material is said to be homogeneous if its constitutive parameters do not vary from point to
point, and isotropic if they are independent of direction. Most
materials are isotropic, but some crystals are not. Throughout
this book, all materials are assumed to be homogeneous
and isotropic. This section is concerned with σ , Section 4-7
examines ε , and discussion of µ is deferred to Chapter 5.
J = σ E (A/m2 )
(Ohm’s law),
(4.73)
where σ is the material’s conductivity with units of siemen per
meter (S/m).
In yet other materials, called dielectrics, the electrons are
tightly bound to the atoms—so much so that it is very difficult
to detach them under the influence of an electric field. Consequently, no significant conduction current can flow through
them.
◮ A perfect dielectric is a material with σ = 0. In
contrast, a perfect conductor is a material with σ = ∞.
Some materials, called superconductors, exhibit such a
behavior. ◭
The conductivity σ of most metals is in the range from 106
to 107 S/m when compared with 10−10 to 10−17 S/m for good
insulators (Table 4-1 on p. 194). A class of materials called
semiconductors allow for conduction currents even though
their conductivities are much smaller than those of metals.
The conductivity of pure germanium, for example, is 2.2 S/m.
Tabulated values of σ at room temperature (20 ◦ C) are given
in Appendix B for some common materials, and a subset is
reproduced in Table 4-1.
◮ The conductivity of a material depends on several factors, including temperature and the presence of impurities.
In general, σ of metals increases with decreasing temperature. Most superconductors operate in the neighborhood
of absolute zero. ◭
4-6 CONDUCTORS
191
Technology Brief 8:
Supercapacitors as Batteries
extending into consumer electronics (cell phones, MP3
players, laptop computers) and electric cars (Fig. TF8-2).
As recent additions to the language of electronics, the
names supercapacitor, ultracapacitor, and nanocapacitor suggest that they represent devices that are
somehow different from or superior to traditional capacitors. Are these just fancy names attached to traditional
capacitors by manufacturers, or are we talking about a
really different type of capacitor?
Capacitor Energy Storage Limitations
◮ The three aforementioned names refer to variations on an energy storage device (Fig. TF8-1)
known by the technical name electrochemical
double-layer capacitor (EDLC), in which energy
storage is realized by a hybrid process that incorporates features from both the traditional electrostatic
capacitor and the electrochemical voltaic battery. ◭
For the purposes of this Technology Brief, we refer
to this relatively new device as a supercapacitor. The
battery is far superior to the traditional capacitor with
regard to energy storage, but a capacitor can be charged
and discharged much more rapidly than a battery. As a
hybrid technology, the supercapacitor offers features that
are intermediate between those of the battery and the
traditional capacitor. The supercapacitor is now used to
support a wide range of applications, from motor startups
in large engines (trucks, locomotives, submarines, etc.)
to flash lights in digital cameras, and its use is rapidly
Energy density W ′ is often measured in watts-hours
per kg (Wh/kg), with 1 Wh = 3.6 × 103 joules. Thus, the
energy capacity of a device is normalized to its mass.
For batteries, W ′ extends between about 30 Wh/kg for a
lead-acid battery to as high as 150 Wh/kg for a lithiumion battery. In contrast, W ′ rarely exceeds 0.02 Wh/kg
for a traditional capacitor. Let us examine what limits
the value of W ′ for the capacitor by considering a small
parallel-plate capacitor with plate area A and separation
between plates d. For simplicity, we assign the capacitor
a voltage rating of 1 V (maximum anticipated voltage
across the capacitor). Our goal is to maximize the energy
density W ′ . For a parallel-plate capacitor C = ε A/d, where
ε is the permittivity of the insulating material. Using
Eq. (4.121) leads to
W′ =
1
ε AV 2
W
=
CV 2 =
m
2m
2md
(J/kg),
where m is the mass of the conducting plates and the
insulating material contained in the capacitor. To keep
the analysis simple, we assume that the plates can be
made so thin as to ignore their masses relative to the
mass of the insulating material. If the material’s density
Figure TF8-1 Examples of supercapacitors. (Courtesy of
Vladimir Zhupanenko/Shutterstock.)
Figure TF8-2 Examples of systems that use supercapacitors.
192
CHAPTER 4
is ρ (kg/m3 ), then m = ρ Ad and
W′ =
εV 2
(J/kg).
2ρ d 2
To maximize W ′ , we need to select d to be the smallest
possible, but we also have to be aware of the constraint
associated with dielectric breakdown. To avoid sparking
between the capacitor’s two plates, the electric field
strength should not exceed Eds , which is the dielectric
strength of the insulating material. Among the various
types of materials commonly used in capacitors, mica
has one of the highest values of Eds , nearly 2 × 108 V/m.
Breakdown voltage Vbr is related to Eds by Vbr = Eds d,
so given that the capacitor is to have a voltage rating
of 1 V, let us choose Vbr to be 2 V, thereby allowing
a 50% safety margin. With Vbr = 2 V and Eds = 2 × 108
V/m, it follows that d should not be smaller than 10−8 m,
or 10 nm. For mica, ε ≈ 6ε0 and ρ = 3 × 103 kg/m3 .
Ignoring for the moment the practical issues associated
with building a capacitor with a spacing of only 10 nm
between conductors, the expression for energy density
leads to W ′ ≈ 90 J/kg. Converting W ′ to Wh/kg (by
dividing by 3.6 × 103 J/Wh) gives
W ′ (max) = 2.5 × 10−2
(Wh/kg),
for a traditional capacitor at a voltage rating of 1 V.
◮ The energy storage capacity of a traditional capacitor is about four orders of magnitude smaller
than the energy density capability of a lithium-ion
battery. ◭
Energy Storage Comparison
The table in the upper part of Fig. TF8-3 displays typical
values or ranges of values for each of five attributes commonly used to characterize the performance of energy
storage devices. In addition to the energy density W ′ ,
they include the power density P′ , the charge and
discharge rates, and the number of charge/discharge
cycles that the device can withstand before deteriorating
in performance. For most energy storage devices, the
discharge rate usually is shorter than the charge rate, but
for the purpose of the present discussion, we treat them
as equal. As a first-order approximation, the discharge
rate is related to P′ and W ′ by
T=
W′
.
P′
ELECTROSTATICS
◮ Supercapacitors are capable of storing 100 to
1000 times more energy than a traditional capacitor
but 10 times less than a battery (Fig. TF8-3). On
the other hand, supercapacitors can discharge their
stored energy in a matter of seconds when compared with hours for a battery. ◭
Moreover, the supercapacitor’s cycle life is on the order of 1 million when compared with only 1000 for a
rechargeable battery. Because of these features, the
supercapacitor has greatly expanded the scope and use
of capacitors in electronic circuits and systems.
Future Developments
The upper right-hand corner of the plot in Fig. TF8-3 represents the ideal energy storage device with W ′ ≈ 100–
1000 Wh/kg and P′ ≈ 103 –104 W/kg. The corresponding
discharge rate is T ≈ 10–100 ms. Current research
aims to extend the capabilities of batteries and supercapacitors in the direction of this prized domain of the
energy-power space.
4-6 CONDUCTORS
193
Energy Storage Devices
Feature
Traditional Capacitor
Energy density W ′ (Wh/kg)
Power density P′ (W/kg)
Charge and discharge rate T
Cycle life Nc
∼ 10−2
1,000 to 10,000
10−3 sec
∞
Energy density W′ (Wh/kg)
1000
Supercapacitor
Battery
1 to 10
5 to 150
1,000 to 5,000
10 to 500
∼ 1 sec to 1 min ∼ 1 to 5 hrs
∼ 106
∼ 103
Future
developments
Fuel cells
100
Batteries
10
Supercapacitors
1
0.1
Traditional
capacitors
0.01
10
100
1000
Power density P′ (W/kg)
Figure TF8-3 Comparison of energy storage devices.
10,000
194
CHAPTER 4
Table 4-1 Conductivity of some common
Material
materials at 20 ◦ C.
Conductivity, σ (S/m)
Conductors
Silver
Copper
Gold
Aluminum
Iron
Mercury
Carbon
Semiconductors
Pure germanium
Pure silicon
Insulators
Glass
Paraffin
Mica
Fused quartz
6.2 × 107
5.8 × 107
4.1 × 107
3.5 × 107
107
106
3 × 104
ELECTROSTATICS
over which the applied electric field can accelerate it before
it is stopped by colliding with an atom and then starts accelerating all over again. From Eq. (4.11), the current density in
a medium containing a volume density ρv of charges moving
with velocity u is J = ρv u. In the most general case, the current
density consists of a component Je due to electrons and a
component Jh due to holes. Thus, the total conduction current
density is
J = Je + Jh = ρve ue + ρvh uh
(A/m2 ),
(4.75)
where ρve = −Ne e and ρvh = Nh e with Ne and Nh being the
number of free electrons and the number of free holes per unit
volume and e = 1.6 × 10−19 C is the absolute charge of a single
hole or electron. Use of Eqs. (4.74a) and (4.74b) gives
2.2
4.4 × 10−4
J = (−ρve µe + ρvh µh )E = σ E,
10−12
10−15
10−15
10−17
(4.76)
where the quantity inside the parentheses is defined as the
conductivity of the material, σ . Thus,
σ = −ρve µe + ρvh µh = (Ne µe + Nh µh )e (S/m),
(4.77a)
(semiconductor)
What are the electromagnetic
constitutive parameters of a material?
Concept Question 4-14:
and its unit is siemens per meter (S/m). For a good conductor,
Nh µh ≪ Ne µe , and Eq. (4.77a) reduces to
σ = −ρve µe = Ne µe e
What classifies a material as
a conductor, a semiconductor, or a dielectric? What is a
superconductor?
Concept Question 4-15:
Concept Question 4-16:
What is the conductivity of a
perfect dielectric?
4-6.1 Drift Velocity
(4.77b)
(good conductor)
◮ In view of Eq. (4.76), in a perfect dielectric with σ = 0,
J = 0 regardless of E. Similarly, in a perfect conductor
with σ = ∞, E = J/σ = 0 regardless of J. ◭
That is,
The drift velocity ue of electrons in a conducting material is
related to the externally applied electric field E through
ue = −µe E
(S/m).
(m/s),
(4.74a)
where µe is a material property call the electron mobility with
units of (m2 /V·s). In a semiconductor, current flow is due to
the movement of both electrons and holes, and since holes are
positive-charge carriers, the hole drift velocity uh is in the same
direction as E,
(m/s),
(4.74b)
uh = µh E
where µh is the hole mobility. The mobility accounts for the
effective mass of a charged particle and the average distance
Perfect dielectric:
J = 0,
Perfect conductor:
E = 0.
Because σ is on the order of 106 S/m for most metals, such as
silver, copper, gold, and aluminum (Table 4-1), it is common
practice to treat them as perfect conductors and to set E = 0
inside them.
A perfect conductor is an equipotential medium, meaning
that the electric potential is the same at every point in the
conductor. This property follows from the fact that V21 , which
is the voltage difference between two points in the conductor,
4-6 CONDUCTORS
195
Module 4.1 Fields Due to Charges For any group of point charges, this module calculates and displays the electric field
E and potential V across a 2-D grid. The user can specify the locations, magnitudes and polarities of the charges.
equals the line integral of E between them, as indicated by
Eq. (4.39). Since E = 0 everywhere in the perfect conductor,
the voltage difference V21 = 0. The fact that the conductor is
an equipotential medium, however, does not necessarily imply
that the potential difference between the conductor and some
other conductor is zero. Each conductor is an equipotential
medium, but the presence of different distributions of charges
on their two surfaces can generate a potential difference
between them.
Example 4-10:
Conduction Current in a
Copper Wire
A 2-mm diameter copper wire with conductivity of
5.8 × 107 S/m and electron mobility of 0.0032 (m2 /V·s)
is subjected to an electric field of 20 (mV/m). Find (a) the
volume charge density of the free electrons, (b) the current
density, (c) the current flowing in the wire, (d) the electron
drift velocity, and (e) the volume density of the free electrons.
Solution: (a)
ρve = −
σ
5.8 × 107
=−
= −1.81 × 1010 (C/m3 ).
µe
0.0032
(b)
J = σ E = 5.8 × 107 × 20 × 10−3 = 1.16 × 106 (A/m2 ).
(c)
I = JA
2
πd
π × 4 × 10−6
= 1.16 × 106
= 3.64 A.
=J
4
4
(d)
ue = −µe E = −0.0032 × 20 × 10−3 = −6.4 × 10−5 m/s.
196
CHAPTER 4
The minus sign indicates that ue is in the opposite direction
of E.
(e)
1.81 × 1010
ρve
Ne = −
= 1.13 × 1029 electrons/m3 .
=
e
1.6 × 10−19
Exercise 4-12: Determine the density of free electrons in
aluminum given that its conductivity is 3.5 × 107 (S/m)
and its electron mobility is 0.0015 (m2 /V · s).
Answer: Ne = 1.46 × 1029 electrons/m3 . (See
EM
potential (point 2). The relation between V and Ex is obtained
by applying Eq. (4.39):
V = V1 − V2 = −
Z x1
x2
E · dl = −
I=
Z
A
J · ds =
Z
.)
To demonstrate the utility of the point form of Ohm’s law,
we apply it to derive an expression for the resistance R of a
conductor of length l and uniform cross section A, as shown
in Fig. 4-16. The conductor axis is along the x direction and
extends between points x1 and x2 , with l = x2 − x1 . A voltage V
applied across the conductor terminals establishes an electric
field E = x̂Ex ; the direction of E is from the point with
higher potential (point 1 in Fig. 4-16) to the point with lower
x
x2
J
E
2 I
A
–
+
(A).
(4.79)
l
σA
(Ω).
(4.80)
We now generalize our result for R to any resistor of
arbitrary shape by noting that the voltage V across the resistor
is equal to the line integral of E over a path l between two
specified points and the current I is equal to the flux of J
through the surface S of the resistor. Thus,
Z
Z
E · dl
V
=Z l
= Z l
.
I
J · ds
σ E · ds
−
E · dl
S
−
(4.81)
S
The reciprocal of R is called the conductance G, and the unit
of G is (Ω−1 ) or siemens (S). For the linear resistor,
G=
1
σA
=
R
l
(S).
(4.82)
Example 4-11: Conductance of Coaxial Cable
y
I 1
(V).
From R = V /I, the ratio of Eq. (4.78) to Eq. (4.79) gives
R=
l
x̂Ex · x̂ dl = Ex l
σ E · ds = σ Ex A
A
R=
4-6.2 Resistance
x1
x2
.)
conducting wire of uniform cross section has a density of
3 × 105 (A/m2 ). Find the voltage drop along the length
of the wire if the wire material has a conductivity of
2 × 107 (S/m).
EM
Z x1
(4.78)
Using Eq. (4.73), the current flowing through the cross
section A at x2 is
Exercise 4-13: The current flowing through a 100-m long
Answer: V = 1.5 V. (See
ELECTROSTATICS
The radii of the inner and outer conductors of a coaxial cable
of length l are a and b, respectively (Fig. 4-17). The insulation
material has conductivity σ . Obtain an expression for G′ ,
which is the conductance per unit length of the insulation layer.
Solution: Let I be the total current flowing radially (along r̂)
from the inner conductor to the outer conductor through the
insulation material. At any radial distance r from the axis of
the center conductor, the area through which the current flows
is A = 2π rl. Hence,
V
Figure 4-16 Linear resistor of cross section A and length l
connected to a dc voltage source V .
J = r̂
I
I
= r̂
,
A
2π rl
(4.83)
I
.
2πσ rl
(4.84)
and from J = σ E,
E = r̂
4-7 DIELECTRICS
197
in the last step of the derivation leading to Eq. (4.88). For a
volume υ , the total dissipated power is
l
E
–
r
Vab
+
a
b
P=
Z
υ
E · J dυ
P=
Figure 4-17 Coaxial cable of Example 4-11.
In a resistor, the current flows from higher electric potential
to lower potential. Hence, if J is in the r̂ direction, the inner
conductor must be at a potential higher than that at the outer
conductor. Accordingly, the voltage difference between the
conductors is
Z a
Z a
I r̂ · r̂ dr
I
b
E · dl = −
Vab = −
.
=
ln
r
2πσ l
a
b 2πσ l
b
(4.85)
The conductance per unit length is then
1
I
2πσ
G
=
=
=
l
Rl Vab l
ln(b/a)
(4.89)
and in view of Eq. (4.73),
σ
G′ =
(W) (Joule’s law),
(S/m).
Z
υ
σ |E|2 d υ
(W).
(4.90)
Equation (4.89) is a mathematical statement of Joule’s law.
For the resistor example considered earlier, |E| = Ex and its
volume is υ = lA. Separating the volume integral in Eq. (4.90)
into a product of a surface integral over A and a line integral
over l, we have
P=
Z
υ
σ |E|2 d υ =
Z
A
σ Ex ds
Z
l
Ex dl
= (σ Ex A)(Ex l) = IV
(W),
(4.91)
where use was made of Eq. (4.78) for the voltage V and
Eq. (4.79) for the current I. With V = IR, we obtain the familiar
expression
P = I2R
(W).
(4.92)
(4.86)
What is the fundamental difference between an insulator, a semiconductor, and a
conductor?
Concept Question 4-17:
4-6.3 Joule’s Law
We now consider the power dissipated in a conducting medium
in the presence of an electrostatic field E. The medium
contains free electrons and holes with volume charge densities ρve and ρvh , respectively. The electron and hole charge
contained in an elemental volume ∆υ is qe = ρve ∆υ and
qh = ρvh ∆υ , respectively. The electric forces acting on qe
and qh are Fe = qe E = ρve E ∆υ and Fh = qh E = ρvh E ∆υ .
The work (energy) expended by the electric field in moving qe
a differential distance ∆le and moving qh a distance ∆lh is
∆W = Fe · ∆le + Fh · ∆lh .
(4.87)
Power P is measured in watts (W) and is defined as the time
rate of change of energy. The power corresponding to ∆W is
∆P =
∆W
∆le
∆lh
= Fe ·
+ Fh ·
∆t
∆t
∆t
= Fe · ue + Fh · uh
= (ρve E · ue + ρvh E · uh ) ∆υ = E · J ∆υ , (4.88)
where ue = ∆le /∆t and uh = ∆lh /∆t are the electron and
hole drift velocities, respectively. Equation (4.75) was used
Show that the power dissipated in the coaxial cable of Fig. 4-17 is
Concept Question 4-18:
P=
I 2 ln(b/a)
.
2πσ l
Exercise 4-14: A 50-m long copper wire has a circular
cross section with radius r = 2 cm. Given that the conductivity of copper is 5.8 × 107 S/m, determine (a) the
resistance R of the wire and (b) the power dissipated in
the wire if the voltage across its length is 1.5 mV.
Answer: (a) R = 6.9 × 10−4 Ω, (b) P = 3.3 mW. (See
EM
.)
Exercise 4-15: Repeat part (b) of Exercise 4-14 by applying Eq. (4.90). (See EM .)
4-7 Dielectrics
The fundamental difference between a conductor and a dielectric is that electrons in the outermost atomic shells of
198
CHAPTER 4
–
–
–
–
Atom
–
–
– –
Electron
E
–
+ + +
E
–
–
– – –
–
– –
+ + +
+
+ +
– – – – – –+ –+ – – – – –+ –+
+ +
+ + + + + + + –
– +
–
+ + – – – –
–
– + +
– – + + + –
+ + + + + + + – –
–
–
– – – – – – – –
+ +
+
– – + + + + + + +
– – – – – – – + + –+ –
– –
+ +
– – –+ + + + + + + + + + –+
– – – – – – – – –
Nucleus
E
Polarized molecule
E
E
Nucleus
(a) External Eext = 0
–
Positive surface charge
E
E
ELECTROSTATICS
q
d
–q
Center of electron cloud
(b) External Eext ≠ 0
Negative surface charge
(c) Electric dipole
Figure 4-18 In the absence of an external electric field E, the
center of the electron cloud is co-located with the center of the
nucleus, but when a field is applied, the two centers are separated
by a distance d.
a conductor are only weakly tied to atoms and hence can
freely migrate through the material, whereas in a dielectric
they are strongly bound to the atom. In the absence of an
electric field, the electrons in nonpolar molecules form a
symmetrical cloud around the nucleus, with the center of the
cloud coinciding with the nucleus (Fig. 4-18(a)). The electric
field generated by the positively charged nucleus attracts and
holds the electron cloud around it, and the mutual repulsion of
the electron clouds of adjacent atoms shapes its form. When
a conductor is subjected to an externally applied electric field,
the most loosely bound electrons in each atom can jump from
one atom to the next, thereby setting up an electric current.
In a dielectric, however, an externally applied electric field E
cannot effect mass migration of charges since none are able to
move freely. Instead, E will polarize the atoms or molecules
in the material by moving the center of the electron cloud
away from the nucleus (Fig. 4-18(b)). The polarized atom or
molecule may be represented by an electric dipole consisting
of charges +q in the nucleus and −q at the center of the
electron cloud (Fig. 4-18(c)). Each such dipole sets up a small
electric field pointing from the positively charged nucleus
to the center of the equally but negatively charged electron
cloud. This induced electric field, called a polarization field,
generally is weaker than and opposite in direction to E.
Consequently, the net electric field present in the dielectric
material is smaller than E. At the microscopic level, each
Figure 4-19 A dielectric medium polarized by an external
electric field E.
dipole exhibits a dipole moment similar to that described in
Example 4-9. Within a block of dielectric material subject to
a uniform external field, the dipoles align themselves linearly,
as shown in Fig. 4-19. Along the upper and lower edges of the
material, the dipole arrangement exhibits positive and negative
surface charge densities, respectively.
It is important to stress that this description applies to
only nonpolar molecules that do not have permanent dipole
moments. Nonpolar molecules become polarized only when
an external electric field is applied; when the field is removed,
the molecules return to their original unpolarized state.
In polar materials such as water, the molecules possess builtin permanent dipole moments that are randomly oriented in
the absence of an applied electric field, and owing to their
random orientations, the dipoles of polar materials produce
no net macroscopic dipole moment (at the macroscopic scale,
each point in the material represents a small volume containing
thousands of molecules). Under the influence of an applied
field, the permanent dipoles tend to align themselves along the
direction of the electric field in a manner similar to that shown
in Fig. 4-19 for nonpolar materials.
4-7.1
Polarization Field
In free space D = ε0 E, the presence of microscopic dipoles in
a dielectric material alters that relationship to
D = ε0 E + P,
(4.93)
4-7 DIELECTRICS
199
where P, called the electric polarization field, accounts for
the polarization properties of the material. The polarization
field is produced by the electric field E and depends on the
material properties. A dielectric medium is said to be linear
if the magnitude of the induced polarization field P is directly
proportional to the magnitude of E, and isotropic if P and E are
in the same direction. Some crystals allow more polarization
to take place along certain directions, such as the crystal
axes, than along others. In such anisotropic dielectrics, E
and P may have different directions. A medium is said to be
homogeneous if its constitutive parameters (ε , µ , and σ ) are
constant throughout the medium. Our present treatment will be
limited to media that are linear, isotropic, and homogeneous.
For such media, P is directly proportional to E and is expressed
as
(4.94)
P = ε0 χe E,
where χe is called the electric susceptibility of the material.
Inserting Eq. (4.94) into Eq. (4.93), we have
D = ε0 E + ε0 χe E = ε0 (1 + χe)E = ε E,
(4.95)
which defines the permittivity ε of the material as
ε = ε0 (1 + χe).
(4.96)
It is often convenient to characterize the permittivity of a
material relative to that of free space, ε0 ; this is accommodated
by the relative permittivity εr = ε /ε0 . Values of εr are listed in
Table 4-2 for a few common materials, and a longer list is
given in Appendix B. In free space εr = 1, and for most conductors, εr ≈ 1. The dielectric constant of air is approximately
1.0006 at sea level and decreases toward unity with increasing
altitude. Except in some special circumstances, such as when
calculating electromagnetic wave refraction (bending) through
the atmosphere over long distances, air can be treated as if it
were free space.
4-7.2 Dielectric Breakdown
The preceding dielectric–polarization model presumes that the
magnitude of E does not exceed a certain critical value, which
is known as the dielectric strength Eds of the material. Beyond
this, electrons will detach from the molecules and accelerate
through the material in the form of a conduction current. When
this happens, sparking can occur, and the dielectric material
can sustain permanent damage due to electron collisions with
the molecular structure. This abrupt change in behavior is
called dielectric breakdown.
◮ The dielectric strength Eds is the largest magnitude of E
that the material can sustain without breakdown. ◭
Dielectric breakdown can occur in gases, liquids, and solids.
The dielectric strength Eds depends on the material composition, as well as other factors such as temperature and humidity.
For air, Eds is roughly 3 (MV/m); for glass, 25 to 40 (MV/m);
and for mica, 200 (MV/m) (see Table 4-2).
A charged thundercloud at electric potential V relative to the
ground induces an electric field E = V/d in the air beneath it,
where d is the height of the cloud base above the ground. If
V is sufficiently large so that E exceeds the dielectric strength
of air, ionization occurs and a lightning discharge follows. The
breakdown voltage Vbr of a parallel-plate capacitor is discussed
in Example 4-12.
Example 4-12: Dielectric Breakdown
In a parallel-plate capacitor with a separation d between the
conducting plates, the electric field E in the dielectric material
Table 4-2 Relative permittivity (dielectric constant) and dielectric strength of common materials.
Material
Air (at sea level)
Petroleum oil
Polystyrene
Glass
Quartz
Bakelite
Mica
Relative Permittivity, εr
Dielectric Strength, Eds (MV/m)
1.0006
2.1
2.6
4.5–10
3.8–5
5
5.4–6
3
12
20
25–40
30
20
200
Note: ε = εr ε0 and ε0 = 8.854 × 10−12 F/m.
200
CHAPTER 4
separating the two plates is related to the voltage V between
the two plates by
V
E= .
d
The breakdown voltage Vbr corresponds to the value of V at
which E = Eds , where Eds is the dielectric strength of the
material contained between the plates. That is,
Vbr = Eds d.
If V exceeds Vbr , the electric charges will “spark” their way
between the two plates.
A thin capacitor filled with quartz operates at 60 V. If
d = 0.01 mm, what is the breakdown voltage, and how does
it compare with the operating voltage?
Solution: From Table 4-2, Eds = 30 × 106 V/m for quartz.
Hence, the breakdown voltage is
Vbr = Eds d = 30 × 106 × 10−5 = 300 V,
which is much higher than the operating voltage of 60 V.
Therefore, the capacitor should experience no issues with
dielectric breakdown.
Concept Question 4-19:
What is a polar material? A
nonpolar material?
Concept Question 4-20: Do D and E always point in
the same direction? If not, when do they not?
Concept Question 4-21:
4-8 Electric Boundary Conditions
◮ A vector field is said to be spatially continuous if it
does not exhibit abrupt changes in either magnitude or
direction as a function of position. ◭
Even though the electric field may be continuous in adjoining dissimilar media, it may well be discontinuous at the
boundary between them. Boundary conditions specify how the
components of fields tangential and normal to an interface
between two media relate across the interface Here we derive
a general set of boundary conditions for E, D, and J that is
applicable at the interface between any two dissimilar media—
be they two dielectrics or a conductor and a dielectric. Of
course, any of the dielectrics may be free space. Even though
these boundary conditions are derived assuming electrostatic
conditions, they remain valid for time-varying electric fields
as well. Figure 4-20 shows an interface between medium 1
with permittivity ε1 and medium 2 with permittivity ε2 . In
the general case, the interface may contain a surface charge
density ρs (unrelated to the dielectric polarization charge
density).
To derive the boundary conditions for the tangential components of E and D, we consider the closed rectangular loop
abcda shown in Fig. 4-20 and apply the conservative property
of the electric field expressed by Eq. (4.40), which states that
the line integral of the electrostatic field around a closed path
is always zero. By letting ∆h → 0, the contributions to the line
integral by segments bc and da vanish. Hence,
What happens when dielectric
Z
breakdown occurs?
C
E1n
E1
b
E1t
E2n
E2t
E2
l̂l1
l̂l2
c
Δl
ELECTROSTATICS
E · dl =
Z b
a
E1 · ℓ̂ℓ1 dl +
}
}
d
Δh
2
Δh
2
Medium 2
ε2
E2 · ℓ̂ℓ2 dl = 0,
nˆ 2
Δs
{
{
Δh
2
Δh
2
c
D1n
Medium 1
ε1
a
Z d
D2n nˆ 1
Figure 4-20 Interface between two dielectric media.
ρs
(4.97)
4-8 ELECTRIC BOUNDARY CONDITIONS
201
where ℓ̂ℓ1 and ℓ̂ℓ2 are unit vectors along segments ab and cd and
E1 and E2 are the electric fields in media 1 and 2. Next, we
decompose E1 and E2 into components tangential and normal
to the boundary (Fig. 4-20),
E1 = E1t + E1n ,
(4.98a)
E2 = E2t + E2n .
(4.98b)
n̂2 ·(D1 − D2) = ρs
(C/m2 ).
(4.103)
If D1n and D2n denote as the normal components of D1 and D2
along n̂2 , we have
Noting that ℓ̂ℓ1 = −ℓ̂ℓ2 , it follows that
(E1 − E2 ) · ℓ̂ℓ1 = 0.
(4.99)
In other words, the component of E1 along ℓ̂ℓ1 equals that of E2
along ℓ̂ℓ1 , for all ℓ̂ℓ1 tangential to the boundary. Hence,
E1t = E2t
medium is always defined to be in the outward direction away
from that medium. Since n̂1 = −n̂2 , Eq. (4.102) simplifies to
(V/m).
(4.100)
◮ Thus, the tangential component of the electric field
is continuous across the boundary between any two
media. ◭
(C/m2 ).
D1 n − D2 n = ρ s
(4.104)
◮ The normal component of D changes abruptly at a
charged boundary between two different media in an
amount equal to the surface charge density. If no charge
exists at the boundary, then Dn is continuous across the
boundary. ◭
The corresponding boundary condition for E is
Upon decomposing D1 and D2 into tangential and normal
components (in the manner of Eq. (4.98)) and noting that
D1t = ε1 E1t and D2t = ε2 E2t , the boundary condition on the
tangential component of the electric flux density is
S
Z
D1 · n̂2 ds +
top
= ρs ∆s,
ε1 E1n − ε2 E2n = ρs .
(4.105b)
(4.101)
Next, we apply Gauss’s law, as expressed by Eq. (4.29),
to determine boundary conditions on the normal components
of E and D. According to Gauss’s law, the total outward flux
of D through the three surfaces of the small cylinder shown in
Fig. 4-20 must equal the total charge enclosed in the cylinder.
By letting the cylinder’s height ∆h → 0, the contribution to the
total flux through the side surface goes to zero. Also, even if
each of the two media happens to contain free charge densities,
the only charge remaining in the collapsed cylinder is that
distributed on the boundary. Thus, Q = ρs ∆s, and
D · ds =
(4.105a)
or equivalently
D1 t
D2 t
=
.
ε1
ε2
Z
n̂2 ·(ε1 E1 − ε2 E2 ) = ρs ,
Z
D2 · n̂1 ds
In summary, (1) the conservative property of E,
C
E · dl = 0,
(4.106)
led to the result that E has a continuous tangential component
across a boundary, and (2) the divergence property of D,
∇ · D = ρv
bottom
Z
×E = 0
∇×
Z
S
D · ds = Q,
(4.107)
(4.102)
where n̂1 and n̂2 are the outward normal unit vectors of
the bottom and top surfaces, respectively. It is important to
remember that the normal unit vector at the surface of any
led to the result that the normal component of D changes by ρs
across the boundary. A summary of the conditions that apply
at the boundary between different types of media is given in
Table 4-3.
202
CHAPTER 4
ELECTROSTATICS
Table 4-3 Boundary conditions for the electric fields.
Field Component
Any Two Media
Medium 1
Dielectric ε1
Medium 2
Conductor
Tangential E
E1t = E2t
E1t = E2t = 0
Tangential D
D1t /ε1 = D2t /ε2
D1t = D2t = 0
Normal E
ε1 E1n − ε2 E2n = ρs
E1n = ρs /ε1
E2n = 0
Normal D
D1n − D2n = ρs
D1n = ρs
D2n = 0
Notes: (1) ρs is the surface charge density at the boundary; (2) normal components of
E1 , D1 , E2 , and D2 are along n̂2 , which is the outward normal unit vector of medium 2.
Example 4-13:
Application of Boundary
Conditions
The x–y plane is a charge-free boundary separating two dielectric media with permittivities ε1 and ε2 , as shown in Fig. 4-21.
If the electric field in medium 1 is
of E1 . The normal to the boundary is ẑ. Hence, the x and y
components of the fields are tangential to the boundary and
the z components are normal to the boundary. At a chargefree interface, the tangential components of E and the normal
components of D are continuous. Consequently,
E2x = E1x ,
E2y = E1y ,
and
E1x + ŷE1y + ẑE1z ,
and E1 = x̂, find (a) the electric field E2 in medium 2 and
(b) the angles θ1 and θ2 .
D2z = D1z
or
ε2 E2z = ε1 E1z .
Hence,
E2 = x̂E1x + ŷE1y + ẑ
z
ε1
E1z .
ε2
(4.108)
(b) The tangential components of E1 and E2 are
E1
E1z
θ1
ε1
x–y plane
E1t
E2z
E2
E1t =
q
2 + E2
E1x
1y
tan θ1 =
E1t
=
E1z
q
2 + E2
E1x
1y
E2t
=
E2z
q
2 + E2
E2x
2y
E2t
Figure 4-21 Application of boundary conditions at the inter-
E2t =
q
2 + E2 .
E2x
2y
The angles θ1 and θ2 are then given by
θ2
ε2
and
tan θ2 =
E1z
E2z
,
=
q
2 + E2
E1x
1y
(ε1 /ε2 )E1z
,
face between two dielectric media (Example 4-13).
and the two angles are related by
Solution: (a) Let E2 = x̂E2x + ŷE2y + ẑE2z . Our task is to
find the components of E2 in terms of the given components
ε2
tan θ2
= .
tan θ1
ε1
(4.109)
4-8 ELECTRIC BOUNDARY CONDITIONS
203
Technology Brief 9:
Capacitive Sensors
To capacitive bridge circuit
To sense is to respond to a stimulus. (See Technology
Brief 7 on resistive sensors.) A capacitor can function
as a sensor if the stimulus changes the capacitor’s
geometry—usually the spacing between its conductive
elements—or the effective dielectric properties of the
insulating material situated between them. Capacitive
sensors are used in a multitude of applications. A few
examples follow.
Fluid Gauge
C
Air
h − hf
w
Fluid
The two metal electrodes in (Fig. TF9-1(a)), usually rods
or plates, form a capacitor whose capacitance is directly
proportional to the permittivity of the material between
them. If the fluid section is of height hf and the height
of the empty space above it is (h − hf ), then the overall
capacitance is equivalent to two capacitors in parallel, or
C = Cf + Ca = εf w
hf
Tank
d
hf
(h − hf)
+ εa w
,
d
d
where w is the electrode plate width, d is the spacing
between electrodes, and εf and εa are the permittivities of
the fluid and air, respectively. Rearranging the expression
as a linear equation yields
C = khf + C0 ,
where the constant coefficient is k = (εf − εa )w/d and
C0 = εa wh/d is the capacitance of the tank when totally
empty. Using the linear equation, the fluid height can
be determined by measuring C with a bridge circuit
(Fig. TF9-1(b)).
◮ The output voltage Vout assumes a functional
form that depends on the source voltage υg , the
capacitance C0 of the empty tank, and the unknown
fluid height hf . ◭
(a) Fluid tank
C0 (empty tank)
R
υg
R
C
Vout
(b) Bridge circuit with 150 kHz ac source
Figure TF9-1 Fluid gauge and associated bridge circuit with
C0 being the capacitance that an empty tank would have and C
the capacitance of the tank under test.
Humidity Sensor
Pressure Sensor
Thin-film metal electrodes shaped in an interdigitized
pattern (to enhance the ratio A/d) are fabricated on a silicon substrate (Fig. TF9-2). The spacing between digits is
typically on the order of 0.2 µ m. The effective permittivity
of the material separating the electrodes varies with the
relative humidity of the surrounding environment. Hence,
the capacitor becomes a humidity sensor.
A flexible metal diaphragm separates an oil-filled chamber with reference pressure P0 from a second chamber
exposed to the gas or fluid whose pressure P is to be
measured by the sensor (Fig. TF9-3(a)). The membrane
is sandwiched—but electrically isolated—between two
conductive parallel surfaces, forming two capacitors in
series Fig. TF9-3(b). When P > P0 , the membrane bends
204
CHAPTER 4
Silicon substrate
ELECTROSTATICS
Electrodes
Fluid
Conducting
plate
1
Figure TF9-2 Interdigital capacitor used as a humidity sensor.
in the direction of the lower plate. Consequently, d1
increases and d2 decreases, and in turn, C1 decreases
and C2 increases (Fig. TF9-3(c)). The converse happens
when P < P0 . With the use of a capacitance bridge
circuit, such as the one in Fig. TF9-1(b), the sensor
can be calibrated to measure the pressure P with good
precision.
Flexible
metallic
membrane
C1
d1
P
2
Oil
P0
C2
d2
3
Conducting
plate
(a) Pressure sensor
Noncontact Sensors
Precision positioning is a critical ingredient in semiconductor device fabrication, as well as in the operation
and control of many mechanical systems. Noncontact
capacitive sensors are used to sense the position of
silicon wafers during the deposition, etching, and cutting
processes, without coming in direct contact with the
wafers.
To bridge circuit
Plate
d1
Membrane
d2
Plate
1 1
2 2
C1
C2
3 3
P = P0
C1 = C2
(b) C1 = C2
◮ Noncontact sensors are also used to sense and
control robot arms in equipment manufacturing and
to position hard disc drives, photocopier rollers, printing presses, and other similar systems. ◭
The concentric plate capacitor in Fig. TF9-4 consists of
two metal plates sharing the same plane but electrically
isolated from each other by an insulating material. When
connected to a voltage source, charges of opposite
polarity form on the two plates, resulting in the creation
of electric field lines between them. The same principle
applies to the adjacent plate’s capacitor in Fig. TF9-5. In
both cases, the capacitance is determined by the shapes
and sizes of the conductive elements and by the effective
permittivity of the dielectric medium containing the electric field lines between them. Often, the capacitor surface
is covered by a thin film of nonconductive material, the
Plate
Membrane
d1 P
Plate
d2
P > P0
To bridge circuit
1 1
C1
2 2
C2
3 3
C1 < C2
(c) C1 < C2
Figure TF9-3
Pressure sensor responds to deflection of
metallic membrane.
4-8 ELECTRIC BOUNDARY CONDITIONS
Conductive plates
205
Electric field lines
C
Insulator
Figure TF9-4 Concentric-plate capacitor.
External object
C0
(a) Adjacent-plates
capacitor
C ≠ C0
(b) Perturbation
field
Figure TF9-5 (a) Adjacent-plates capacitor; (b) perturbation
field.
purpose of which is to keep the plate surfaces clean and
dust free.
◮ The introduction of an external object into the
proximity of the capacitor (Fig. TF9-5(b)) changes
the effective permittivity of the medium, perturbs the
electric field lines, and modifies the charge distribution on the plates. ◭
This, in turn, changes the value of the capacitance as it
would be measured by a capacitance meter or bridge
circuit. Hence, the capacitor becomes a proximity
sensor, and its sensitivity depends, in part, on how
different the permittivity of the external object is from that
of the unperturbed medium and on whether it is or is not
made of a conductive material.
Fingerprint Imager
An interesting extension of noncontact capacitive sensors is the development of a fingerprint imager consisting
of a two-dimensional array of capacitive sensor cells
constructed to record an electrical representation of a
fingerprint (Fig. TF9-6). Each sensor cell is composed of
an adjacent-plates capacitor connected to a capacitance
measurement circuit (Fig. TF9-7). The entire surface of
the imager is covered by a thin layer of nonconductive
oxide. When the finger is placed on the oxide surface,
it perturbs the field lines of the individual sensor cells to
varying degrees, depending on the distance between the
ridges and valleys of the finger’s surface from the sensor
cells.
◮ Given that the dimensions of an individual sensor
are on the order of 65 µ m on the side, the imager
is capable of recording a fingerprint image at a
resolution corresponding to 400 dots per inch or
better. ◭
206
CHAPTER 4
Figure TF9-6 Elements of a fingerprint matching system. (Courtesy of IEEE Spectrum.)
Figure TF9-7 Fingerprint representation.
ELECTROSTATICS
4-8 ELECTRIC BOUNDARY CONDITIONS
207
These two boundary conditions can be combined into
Exercise 4-16: Find E1 in Fig. 4-21 if
E2 = x̂2 − ŷ3 + ẑ3 (V/m),
D1 = ε1 E1 = n̂ρs ,
ε1 = 2ε0 ,
(4.111)
(at conductor surface)
ε2 = 8ε0 ,
where n̂ is a unit vector directed normally outward from the
conducting surface.
and the boundary is charge-free.
Answer: E1 = x̂2 − ŷ3 + ẑ12 (V/m). (See
EM
.)
Exercise 4-17: Repeat Exercise 4.16 for a boundary with
◮ The electric field lines point directly away from the
conductor surface when ρs is positive and directly toward
the conductor surface when ρs is negative. ◭
surface charge density ρs = 3.54 × 10−11 (C/m2 ).
Answer: E1 = x̂2 − ŷ3 + ẑ14 (V/m). (See
EM
.)
Figure 4-22 shows an infinitely long conducting slab placed
in a uniform electric field E1 . The media above and below the
slab have permittivity ε1 . Because E1 points away from the
upper surface, it induces a positive charge density ρs = ε1 |E1 |
on the upper slab surface. On the bottom surface, E1 points
toward the surface; therefore, the induced charge density is
−ρs. The presence of these surface charges induces an electric
field Ei in the conductor, resulting in a total field E = E1 + Ei .
To satisfy the condition that E must be everywhere zero in the
conductor, Ei must equal −E1 .
If we place a metallic sphere in an electrostatic field
(Fig. 4-23), positive and negative charges accumulate on the
upper and lower hemispheres, respectively. The presence of
the sphere causes the field lines to bend to satisfy the condition
expressed by Eq. (4.111); that is, E is always normal to a
conductor boundary.
4-8.1 Dielectric-Conductor Boundary
Consider the case when medium 1 is a dielectric and medium 2
is a perfect conductor. In a perfect conductor, because electric
fields and fluxes vanish, it follows that E2 = D2 = 0, which
implies that components of E2 and D2 tangential and normal
to the interface are zero. Consequently, from Eq. (4.100) and
Eq. (4.104), the fields in the dielectric medium at the boundary
with the conductor satisfy
E1t = D1t = 0,
(4.110a)
D1n = ε1 E1n = ρs .
(4.110b)
E1
+
+
+
E1
+
E1
Conducting slab
–
–
–
+
+
+
Ei
–
+
+
E1
–
–
–
–
E1
ε1
+
+
+
Ei
–
+
ρs = ε1E1
+
E1
–
–
ε1
–
–
+
+
Ei
–
–
–
−ρs
Figure 4-22 When a conducting slab is placed in an external electric field E1 , charges that accumulate on the conductor surfaces induce
an internal electric field Ei = −E1 . Consequently, the total field inside the conductor is zero.
208
CHAPTER 4
ELECTROSTATICS
Module 4.2 Charges in Adjacent Dielectrics In two adjoining half-planes with selectable permittivities, the user can
place point charges anywhere in space and select their magnitudes and polarities. The module then displays E, V , and the
equipotential contours of V .
4-8.2
E0
+
+ + + +
+
+
metal
sphere
–
–
––
Figure 4-23
field E0 .
+
+
+
–
–
–
– – –
Metal sphere placed in an external electric
Conductor–Conductor Boundary
We now examine the general case of the boundary between
two media—neither of which is a perfect dielectric or a perfect
conductor (Fig. 4-24). Medium 1 has permittivity ε1 and
conductivity σ1 , medium 2 has ε2 and σ2 , and the interface
between them holds a surface charge density ρs . For the
electric fields, Eqs. (4.100) and (4.105b) give
E1t = E2t ,
ε1 E1n − ε2 E2n = ρs .
(4.112)
Since we are dealing with conducting media, the electric fields
give rise to current densities J1 = σ1 E1 and J2 = σ2 E2 . Hence,
J 2t
J 1t
=
,
σ1
σ2
ε1
J1n
J2n
− ε2
= ρs .
σ1
σ2
(4.113)
The tangential current components J1t and J2t represent currents flowing in the two media in a direction parallel to the
4-9 CAPACITANCE
209
J1n
J1
Medium 1
ε1, σ1
+
n̂
+
J1t
Medium 2
ε2, σ2
J2t
J2n
V
J2
Figure 4-24 Boundary between two conducting media.
boundary; hence, there is no transfer of charge between them.
This is not the case for the normal components. If J1n 6= J2n ,
then a different amount of charge arrives at the boundary than
leaves it. Hence, ρs cannot remain constant in time, which
violates the condition of electrostatics requiring all fields
and charges to remain constant. Consequently, the normal
component of J has to be continuous across the boundary
between two different media under electrostatic conditions.
Upon setting J1n = J2n in Eq. (4.113), we have
J1n
ε1
ε2
−
σ1 σ2
= ρs
(electrostatics).
(4.114)
What are the boundary conditions for the electric field at a conductor–dielectric
boundary?
Concept Question 4-22:
Concept Question 4-23:
Under electrostatic conditions, we require J1n = J2n
at the boundary between two conductors. Why?
4-9 Capacitance
When separated by an insulating (dielectric) medium, any two
conducting bodies, regardless of their shapes and sizes, form
a capacitor. If a dc voltage source is connected across them
(Fig. 4-25) the surfaces of the conductors connected to the
positive and negative source terminals accumulate charges +Q
and −Q, respectively.
+
+
+
+Q
Conductor 1
+
+ +
+
Surface S
+
+
ρs
+
E
–
–
– –
–
–
−Q
Conductor 2 –
–
– –
– –
Figure 4-25 A dc voltage source connected to a capacitor
composed of two conducting bodies.
◮ When a conductor has excess charge, it distributes the
charge on its surface in such a manner as to maintain
a zero electric field everywhere within the conductor,
thereby ensuring that the electric potential is the same at
every point in the conductor. ◭
The capacitance of a two-conductor configuration is defined
as
Q
(C/V or F),
(4.115)
C=
V
where V is the potential (voltage) difference between the
conductors. Capacitance is measured in farads (F), which is
equivalent to coulombs per volt (C/V).
The presence of free charges on the conductors’ surfaces
gives rise to an electric field E (Fig. 4-25) with field lines
originating on the positive charges and terminating on the
negative ones. Since the tangential component of E always
vanishes at a conductor’s surface, E is always perpendicular
to the conducting surfaces. The normal component of E at any
point on the surface of either conductor is given by
En = n̂ · E =
ρs
,
ε
(4.116)
(at conductor surface)
where ρs is the surface charge density at that point, n̂ is
the outward normal unit vector at the same location, and ε
is the permittivity of the dielectric medium separating the
210
CHAPTER 4
ELECTROSTATICS
Module 4.3 Charges above a Conducting Plane When electric charges are placed in a dielectric medium adjoining
a conducting plane, some of the conductor’s electric charges move to its surface boundary, thereby satisfying the boundary
conditions outlined in Table 4-3. This module displays E and V everywhere and ρs along the dielectric–conductor boundary.
conductors. The charge Q is equal to the integral of ρs over
surface S (Fig. 4-25):
Q=
Z
S
ρs ds =
Z
S
ε n̂ · E ds =
Z
S
ε E · ds,
(4.117)
where use was made of Eq. (4.116). The voltage V is related
to E by Eq. (4.39):
V = V12 = −
Z P1
P2
E · dl,
(4.118)
where points P1 and P2 are any two arbitrary points on
conductors 1 and 2, respectively. Substituting Eqs. (4.117) and
(4.118) into Eq. (4.115) gives
C=
Z
ε E · ds
SZ
−
l
(F),
(4.119)
E · dl
where l is the integration path from conductor 2 to conductor 1.
To avoid making sign errors when applying Eq. (4.119), it is
important to remember that surface S is the +Q surface and
P1 is on S. [Alternatively, if you compute C and it comes out
negative, just change its sign.] Because E appears in both the
numerator and denominator of Eq. (4.119), the value of C
obtained for any specific capacitor configuration is always
independent of E’s magnitude. In fact, C depends only on the
capacitor geometry (sizes, shapes and relative positions of the
two conductors) and the permittivity of the insulating material.
4-9 CAPACITANCE
211
Module 4.4 Charges near a Conducting Sphere This module is similar to Module 4.3, except that now the conducting
body is a sphere of selectable size.
If the material between the conductors is not a perfect
dielectric (i.e., if it has a small conductivity σ ), then current
can flow through the material between the conductors, and the
material exhibits a resistance R. The general expression for R
for a resistor of arbitrary shape is given by Eq. (4.81):
−
R= Z
S
Z
l
E · dl
(Ω).
(4.120)
σ E · ds
For a medium with uniform σ and ε , the product of
Eqs. (4.119) and (4.120) gives
RC =
ε
.
σ
(4.121)
This simple relation allows us to find R if C is known, and vice
versa.
Example 4-14:
Capacitance of Parallel-Plate
Capacitor
Obtain an expression for the capacitance C of a parallel-plate
capacitor comprised of two parallel plates each of surface
area A and separated by a distance d. The capacitor is filled
with a dielectric material with permittivity ε .
Solution: In Fig. 4-26, we place the lower plate of the capacitor in the x–y plane and the upper plate in the plane z = d.
Because of the applied voltage difference V , charges +Q and
−Q accumulate on the top and bottom capacitor plates. If the
plate dimensions are much larger than the separation d, then
these charges distribute themselves quasi-uniformly across
the plates, giving rise to a quasi-uniform field between them
pointing in the −ẑ direction. In addition, a fringing field will
exist near the capacitor edges, but its effects may be ignored
because the bulk of the electric field exists between the plates.
212
CHAPTER 4
z
Conducting plate
Area A
ρs
z=d
+
V
+Q
+ + + + + + + + + +
+
ds
–
z=0
ELECTROSTATICS
E E E
– – – – – – – – – –
–ρs
–
+
+
+
+
–
E –
E –
–
+
–
+
Fringing
field lines
–
Dielectric ε
–Q
Conducting plate
Figure 4-26 A dc voltage source connected to a parallel-plate capacitor (Example 4-14).
The charge density on the upper plate is ρs = Q/A. Hence, in
the dielectric medium
Example 4-15:
Capacitance per Unit Length
of Coaxial Line
E = −ẑE,
and from Eq. (4.116), the magnitude of E at the conductor–
dielectric boundary is E = ρs /ε = Q/ε A. From Eq. (4.118),
the voltage difference is
V =−
Z d
0
E · dl = −
Z d
0
(−ẑE) · ẑ dz = Ed,
(4.122)
and the capacitance is
C=
Q
Q
εA
=
=
,
V
Ed
d
(4.123)
where use was made of the relation E = Q/ε A.
Obtain an expression for the capacitance of the coaxial line
shown in Fig. 4-27.
Solution: For a given voltage V across the capacitor, charges
+Q and −Q accumulate on the surfaces of the outer and inner
conductors, respectively. We assume that these charges are
uniformly distributed along the length and circumference of
the conductors with surface charge density ρs′ = Q/2π bl on
the outer conductor and ρs′′ = −Q/2π al on the inner one.
Ignoring fringing fields near the ends of the coaxial line, we
can construct a cylindrical Gaussian surface in the dielectric in
between the conductors with the radius r such that a < r < b.
Symmetry implies that the E-field is identical at all points on
this surface, which is directed radially inward. From Gauss’s
law, it follows that the field magnitude equals the absolute
l
+
V –
b
a
+ + + + + + + + + + + + +
E
E
E
– – – – – – – – – – – – –
– – – – – – – – – – – – –
E
E
E
+ + + + + + + + + + + + +
ρl
–ρl
Inner conductor
Dielectric material ε
Outer conductor
Figure 4-27 Coaxial capacitor filled with insulating material of permittivity ε (Example 4-15).
4-10 ELECTROSTATIC POTENTIAL ENERGY
213
value of the total charge enclosed, divided by the surface area.
That is,
Q
E = −r̂
.
(4.124)
2πε rl
The potential difference V between the outer and inner conductors is
Z b
Z b
Q
−r̂
E · dl = −
V =−
· (r̂ dr)
2πε rl
a
a
Q
b
.
(4.125)
ln
=
2πε l
a
From the definition of υ , the amount of work dWe required
to transfer an additional incremental charge dq from one
conductor to the other is
dWe = υ dq =
2πε l
Q
=
,
V
ln(b/a)
(4.129)
If we transfer a total charge Q between the conductors of an
initially uncharged capacitor, then the total amount of work
performed is
We =
The capacitance C is then given by
C=
q
dq.
C
Z Q
q
0
C
dq =
1 Q2
2 C
(J).
(4.130)
Using C = Q/V , where V is the final voltage, We also can be
expressed as
(4.126)
We = 21 CV 2
(J).
(4.131)
and the capacitance per unit length of the coaxial line is
C′ =
2πε
C
=
l
ln(b/a)
(F/m).
(4.127)
How is the capacitance of a
two-conductor structure related to the resistance of the
insulating material between the conductors?
Concept Question 4-24:
The capacitance of the parallel-plate capacitor discussed in
Example 4-14 is given by Eq. (4.123) as C = ε A/d, where A is
the surface area of each of its plates and d is the separation
between them. Also, the voltage V across the capacitor is
related to the magnitude of the electric field E in the dielectric
by V = Ed. Using these two expressions in Eq. (4.131) gives
We =
Concept Question 4-25:
What are fringing fields and
when may they be ignored?
4-10 Electrostatic Potential Energy
A source connected to a capacitor expends energy in charging
up the capacitor. If the capacitor plates are made of a good
conductor with effectively zero resistance, and if the dielectric
separating the two plates has negligible conductivity, then no
real current can flow through the dielectric, and no ohmic
losses occur anywhere in the capacitor. Where then does the
energy expended in charging up the capacitor go? The energy
ends up getting stored in the dielectric medium in the form of
electrostatic potential energy. The amount of stored energy We
is related to Q, C, and V .
Suppose we were to charge up a capacitor by ramping up
the voltage across it from υ = 0 to υ = V . During the process,
charge +q accumulates on one conductor and −q on the other.
In effect, a charge q has been transferred from one of the
conductors to the other. The voltage υ across the capacitor is
related to q by
q
υ= .
(4.128)
C
1
2
εA
(Ed)2 =
d
1
2
ε E 2 (Ad) =
1
2
ε E 2υ ,
(4.132)
where υ = Ad is the volume of the capacitor. This expression
affirms the assertion made at the beginning of this section,
namely that the energy expended in charging up the capacitor
is being stored in the electric field present in the dielectric
material in between the two conductors.
The electrostatic energy density we is defined as the electrostatic potential energy We per unit volume:
we =
We
υ
=
1
εE2
2
(J/m3 ).
(4.133)
Even though this expression was derived for a parallel-plate
capacitor, it is equally valid for any dielectric medium containing an electric field E, including a vacuum. Furthermore, for
any volume υ , the total electrostatic potential energy stored in
it is
Z
1
ε E 2 dυ
(J).
(4.134)
We =
2 υ
Returning to the parallel-plate capacitor, the oppositely
charged plates are attracted to each other by an electrical
force Fe . In terms of the coordinate system of Fig. 4-28, the
214
CHAPTER 4
z
Conducting plate
Area A
z=d
+
V
+ + + + + + + + + +
+
ds
–
z=0
ELECTROSTATICS
E E E
– – – – – – – – – –
–
+
+
+
E –
E –
–
+
–
+
–
+
Fringing
field lines
–
Dielectric ε
Conducting plate
Figure 4-28 A dc voltage source connected to a parallel-plate capacitor.
electrical force acting on the upper plate is along −ẑ (due to
attraction by the lower plate). Hence, it is given by
Fe = −ẑ Fe
(force on upper plate).
(4.135)
Our plan is to compute Fe from energy considerations. We
start by converting the spacing d into a variable z and using
C = ε A/z in Eq. (4.131):
We =
1
2
CV 2 =
1
2
ε AV 2
.
z
(4.136)
If V is maintained at a constant level, We decreases when
increasing the separation z between the plates. If an external,
upward-directed force F = −Fe is applied to counter the
electrostatic force Fe and used to move the upper plate upwards
by a distance dz, the expended mechanical work is
dW = F · ẑ dz.
(4.137)
The work dW is equal to the loss in electrostatic energy stored
in the capacitor. That is,
dW = −dWe .
(4.138)
Also, Fe = −F, which leads to
dWe = Fe · ẑ dz = −ẑ Fe · ẑ dz = −Fe dz.
(4.139)
From Eq. (4.136),
dWe = − 12 ε
AV 2
dz.
z2
(4.140)
Equating Eqs. (4.139) and (4.140) and replacing z with d leads
to
AV 2
Fe = 12 ε 2 ,
(4.141a)
d
and
V2
(N).
d2
(parallel-plate capacitor)
Fe = −ẑ
1
2
εA
(4.141b)
This is the electrostatic force exerted on the upper plate. The
force on the lower plate is identical in magnitude and opposite
in direction.
The relationship given by Eq. (4.139) pertains to a capacitor
with dl = ẑ dz. We can generalize the result for dl along any
direction as
Fe = −∇We .
(4.142)
Example 4-16: Force on Sliding Dielectric
The two plates of the parallel-plate capacitor shown in
Fig. 4-29 are each of length ℓ and width w, and the separation
between them is d. The capacitor contains a dielectric block of
dimensions ℓ × w × d and permittivity ε . The block can slide
in and out of the capacitor cavity along its length dimension.
Compute the force Fe acting on the dielectric block when it
is partially outside of the cavity and the voltage across the
capacitor is V .
Solution: From Eq. (4.122), the electric field inside the
capacitor cavity is
V
E= .
d
This is true in both the section containing the dielectric block
and the section filled with air. The total electrostatic energy of
4-11 IMAGE METHOD
215
x
y
Dielectric
ε
F
Air
ε0
Fe
E
d
V
x
ℓ
Figure 4-29 Parallel-plate capacitor with slidable dielectric block.
the capacitor consists of two components: one for the volume
containing the dielectric block of permittivity ε and volume
υ 1 = xwd, and another for the volume containing air with ε0
and volume υ 2 = (ℓ − x)wd. Hence,
We =
1
2
=
1
2
=
1
2
ε E 2 υ 1 + 12 ε0 E 2 υ 2
2
2
V
V
1
ε
xwd + 2 ε0
(ℓ − x)wd
d
d
V2
w[ε x + ε0 (ℓ − x)].
d
(4.143)
Since ε > ε0 , the electrostatic energy is maximum when x = ℓ
(dielectric block fully inside the cavity). Sliding the dielectric block out of the capacitor requires exerting an external
mechanical force F to oppose the electrostatic force Fe , whose
tendency is to oppose reduction in We . Thus, the direction of
Fe is to pull the block back into the capacitor.
The magnitude of Fe can be obtained from
dWe
dx
d 1 V2
w[ε x + ε0 (ℓ − x)]
=
dx 2 d
Fe =
=
1 V2
w(ε − ε0 ).
2 d
(4.144)
To bring a charge q from
infinity to a given point in space, a certain amount of
work W is expended. Where does the energy corresponding to W go?
Concept Question 4-26:
When a voltage source is connected across a capacitor, what is the direction of the
electrical force acting on its two conducting surfaces?
Concept Question 4-27:
Exercise 4-18: The radii of the inner and outer conductors
of a coaxial cable are 2 cm and 5 cm, respectively, and
the insulating material between them has a relative permittivity of 4. The charge density on the outer conductor
is ρℓ = 10−4 (C/m). Use the expression for E derived in
Example 4-15 to calculate the total energy stored in a
20 cm length of the cable.
Answer: We = 4.1 J. (See
EM
.)
4-11 Image Method
Consider a point charge Q at a distance d above a horizontally
infinite, perfectly conducting plate (Fig. 4-30(a)). We want to
determine V and E at any point in the space above the plate,
as well as the surface charge distribution on the plate. Three
different methods for finding E have been introduced in this
chapter The first method, based on Coulomb’s law, requires
knowledge of the magnitudes and locations of all the charges.
In the present case, the charge Q induces an unknown and
nonuniform distribution of charge on the plate. Hence, we
cannot utilize Coulomb’s method. The second method, based
on Gauss’s law, is equally difficult to use because it is not
clear how to construct a Gaussian surface across which E is
only tangential or only normal. The third method is based
on evaluating the electric field using E = −∇V after solving
Poisson’s or Laplace’s equation for V subject to the available
boundary conditions, but it is mathematically involved.
Alternatively, the problem at hand can be solved using
image theory.
216
CHAPTER 4
z
+
d
ε
V=0
Electric field
lines
Q
+
d
n̂
d
–
σ=∞
(a) Charge Q above grounded plane
ELECTROSTATICS
Q
ε
V=0
ε
−Q
(b) Equivalent configuration
Figure 4-30 By image theory, a charge Q above a grounded, perfectly conducting plane is equivalent to Q and its image −Q with the
grounded plane removed.
◮ Any given charge configuration above an infinite,
perfectly conducting plane is electrically equivalent to the
combination of the given charge configuration and its image configuration with the conducting plane removed. ◭
The image-method equivalent of the charge Q above a
conducting plane is shown in the right-hand section of
Fig. 4-30. It consists of the charge Q itself and an image
charge −Q at a distance 2d from Q with nothing else between
them. The electric field due to the two isolated charges can now
be easily found at any point (x, y, z) by applying Coulomb’s
method, as demonstrated by Example 4-17. By symmetry, the
combination of the two charges always produces a potential
V = 0 at every point in the plane previously occupied by the
conducting surface. If the charge resides in the presence of
more than one grounded plane, it is necessary to establish its
images relative to each of the planes and then to establish
images of each of those images against the remaining planes.
ρl
ρv
The process is continued until the condition V = 0 is satisfied
everywhere on all grounded planes. The image method applies
not only to point charges but also to distributions of charge,
such as the line and volume distributions depicted in Fig. 4-31.
Once E has been determined, the charge induced on the plate
can be found from
ρs = (n̂ · E)ε0 ,
where n̂ is the normal unit vector to the plate (Fig. 4-30(a)).
Example 4-17:
Image Method for Charge
above Conducting Plane
Use image theory to determine E at an arbitrary point
P = (x, y, z) in the region z > 0 due to a charge Q in free space
at a distance d above a grounded conducting plate residing in
the z = 0 plane.
ρl
ε
ρv
V=0
ε
V=0
ε
σ=∞
–ρl
(a) Charge distributions above grounded plane
(4.145)
–ρv
(b) Equivalent distributions
Figure 4-31 Charge distributions above a conducting plane and their image-method equivalents.
4-11 IMAGE METHOD
217
z
P = (x, y, z)
R1
Q = (0, 0, d) +
R2
z = 0 plane
E=
1
4πε0
QR1 −QR2
+
R31
R32
x̂x + ŷy + ẑ(z + d)
x̂x + ŷy + ẑ(z − d)
Q
−
=
4πε0 [x2 + y2 + (z − d)2]3/2 [x2 + y2 + (z + d)2]3/2
for z ≥ 0.
–Q = (0, 0, –d) –
What is the fundamental
premise of the image method?
Concept Question 4-28:
Figure 4-32 Application of the image method for finding E at
point P (Example 4-17).
Given a charge distribution,
what are the various approaches described in this chapter
for computing the electric field E at a given point in space?
Concept Question 4-29:
Solution: In Fig. 4-32, charge Q is at (0, 0, d), and its
image −Q is at (0, 0, −d). From Eq. (4.19), the electric field at
point P(x, y, z) due to the two charges is given by the following
equation.
Exercise 4-19: Use the result of Example 4-17 to find the
surface charge density ρs on the surface of the conducting
plane.
Answer: ρs = −Qd/[2π (x2 + y2 + d 2 )3/2 ]. (See
EM
.)
Chapter 4 Summary
Concepts
• Maxwell’s equations are the fundamental tenets of
electromagnetic theory.
• Under static conditions, Maxwell’s equations separate
into two uncoupled pairs with one pair pertaining to
electrostatics and the other to magnetostatics.
• Coulomb’s law provides an explicit expression for the
electric field due to a specified charge distribution.
• Gauss’s law states that the total electric field flux
through a closed surface is equal to the net charge
enclosed by the surface.
• The electrostatic field E at a point is related to the
electric potential V at that point by E = −∇V with V
often being referenced to zero at infinity.
• Because most metals have conductivities on the order
of 106 (S/m), they are treated in practice as perfect
conductors. By the same token, insulators with conduc-
•
•
•
•
tivities smaller than 10−10 (S/m) often are treated as
perfect dielectrics.
Boundary conditions at the interface between two materials specify the relations between the normal and
tangential components of D, E, and J in one of the
materials to the corresponding components in the other.
The capacitance of a two-conductor body and resistance of the medium between them can be computed
from knowledge of the electric field in that medium.
The electrostatic energy density stored in a dielectric
medium is we = 21 ε E 2 (J/m3 ).
When a charge configuration exists above an infinite,
perfectly conducting plane, the induced field E is the
same as that due to the configuration itself and its image
with the conducting plane removed.
218
CHAPTER 4
ELECTROSTATICS
Mathematical and Physical Models
Maxwell’s Equations for Electrostatics
Name
Gauss’s law
Kirchhoff’s law
Differential Form
∇ · D = ρv
Integral
Form
Z
×E = 0
∇×
S
Z
D · ds = Q
C
E · dl = 0
Electric Field
Current density
J = ρv u
Poisson’s equation
∇2V = −
Laplace’s equation
Resistance
Boundary conditions
Capacitance
∇2V
ρv
ε
E = R̂
Many point charges
E=
1
4πε
qi (R − Ri )
3
i=1 |R − Ri |
Volume distribution
E=
1
4πε
Z
R̂
Surface distribution
E=
1
4πε
Z
R̂
Line distribution
E=
1
4πε
Z
R̂
Infinite sheet of charge
E = ẑ
Infinite line of charge
E=
ρℓ
D
Dr
= r̂
= r̂
ε0
ε0
2πε0 r
Dipole
E=
qd
(R̂ 2 cos θ + θ̂θ sin θ )
4πε0 R3
Relation to V
E = −∇V
=0
−
R= Z
S
Z
l
E · dl
σ E · ds
Table 4-3
C=
Z
ε E · ds
SZ
−
l
E · dl
ε
σ
RC relation
RC =
Energy density
we = 21 ε E 2
PROBLEMS
Sections 4-2: Charge and Current Distributions
∗ 4.1 A cube 2 m on a side is located in the first octant in
a Cartesian coordinate system with one of its corners at the
origin. Find the total charge contained in the cube if the charge
density is given by ρv = xy2 e−2z (mC/m3 ).
∗ Answer(s) available in Appendix E.
q
4πε R2
Point charge
N
∑
υ′
S′
l′
′
ρv d υ ′
R′2
′
ρs ds′
R′2
′
ρℓ dl ′
R′2
ρs
2ε0
4.2 Find the total charge contained in a cylindrical volume
defined by r ≤ 2 m and 0 ≤ z ≤ 3 m if ρv = 20rz (mC/m3 ).
∗ 4.3 Find the total charge contained in a round-top cone
defined by R ≤ 2 m and 0 ≤ θ ≤ π /4 given that
ρv = 10R2 cos2 θ (mC/m3 ).
4.4 If the line charge density is given by ρl = 24y2 (mC/m),
find the total charge distributed on the y axis from y = −5 to
y = 5.
PROBLEMS
Important Terms
219
Provide definitions or explain the meaning of the following terms:
boundary conditions
capacitance C
charge density
conductance G
conduction current
conductivity σ
conductor
conservative field
constitutive parameters
convection current
Coulomb’s law
current density J
dielectric breakdown voltage Vbr
dielectric material
dielectric strength Eds
dipole moment p
electric dipole
electric field intensity E
electric flux density D
electric potential V
electric susceptibility χe
electron drift velocity ue
electron mobility µe
electrostatic energy density we
electrostatic potential energy We
electrostatics
equipotential
Gaussian surface
Gauss’s law
hole drift velocity uh
hole mobility µh
homogeneous material
image method
isotropic material
4.5 Find the total charge on a circular disk defined by r ≤ a
and z = 0 if:
(a) ρs = ρs0 cos φ (C/m2 )
(b) ρs = ρs0 sin2 φ (C/m2 )
(c) ρs = ρs0 e−r (C/m2 )
(d) ρs = ρs0 e−r sin2 φ (C/m2 )
where ρs0 is a constant.
4.6 If J = ŷ4xz (A/m2 ), find the current I flowing through a
square with corners at (0, 0, 0), (2, 0, 0), (2, 0, 2), and (0, 0, 2).
∗ 4.7 If J = R̂5/R (A/m2 ), find I through the surface R = 5 m.
4.8 An electron beam shaped like a circular cylinder of
radius r0 carries a charge density given by
−ρ0
ρv =
(C/m3 )
1 + r2
where ρ0 is a positive constant and the beam’s axis is coincident with the z axis.
(a) Determine the total charge contained in length L of the
beam.
(b) If the electrons are moving in the +z direction with
uniform speed u, determine the magnitude and direction
of the current crossing the z-plane.
4.9 A circular beam of charge of radius a consists of electrons moving with a constant speed u along the +z direction.
Joule’s law
Kirchhoff’s voltage law
Laplace’s equation
linear material
Ohm’s law
perfect conductor
perfect dielectric
permittivity ε
Poisson’s equation
polarization vector P
relative permittivity εr
semiconductor
static condition
superconductor
volume, surface, and line
charge densities
The beam’s axis is coincident with the z axis and the electron
charge density is given by
ρv = −cr2
(c/m3 )
where c is a constant and r is the radial distance from the axis
of the beam.
∗ (a) Determine the charge density per unit length.
(b) Determine the current crossing the z-plane.
4.10 A line of charge of uniform density ρℓ occupies a
semicircle of radius b as shown in Fig. P4.10. Use the material
presented in Example 4-4 to determine the electric field at the
origin.
z
ρl
y
x
b
Figure P4.10 Problem 4.10.
220
CHAPTER 4
Section 4-3: Coulomb’s Law
∗ 4.11 A square with sides of 2 m has a charge of 40 µ C at
ELECTROSTATICS
shown in Fig. P4.19. If the two right triangles are symmetrical
and of equal corresponding sides, show that the electric field is
zero at the origin.
each of its four corners. Determine the electric field at a point
5 m above the center of the square.
4.12 Three point charges, each with q = 3 nC, are located at
the corners of a triangle in the x–y plane, with one corner at
the origin, another at (2 cm, 0, 0), and the third at (0, 2 cm, 0).
Find the force acting on the charge located at the origin.
y
−2ρl
∗ 4.13 Charge q = 6 µ C is located at (1 cm, 1 cm, 0) and
1
charge q2 is located at (0, 0, 4 cm). What should q2 be so that
E at (0, 2 cm, 0) has no y component?
ρl
ρl
4.14 A line of charge with uniform density ρℓ = 8 (µ C/m)
exists in air along the z axis between z = 0 and z = 5 cm. Find
E at (0,10 cm,0).
4.15 Electric charge is distributed along an arc located in the
x–y plane and defined by r = 2 cm and 0 ≤ φ ≤ π /4. If ρℓ = 5
(µ C/m), find E at (0, 0, z) and then evaluate it at
x
∗ (a) the origin,
(b) z = 5 cm, and
(c) z = −5 cm.
Figure P4.19 Kite-shaped arrangement of line charges for
Problem 4.19.
4.16 A line of charge with uniform density ρl extends
between z = −L/2 and z = L/2 along the z axis. Apply
Coulomb’s law to obtain an expression for the electric field
at any point P(r, φ , 0) on the x–y plane. Show that your result
reduces to the expression given by Eq. (4.33) as the length L is
extended to infinity.
∗ 4.17 Repeat Example 4-5 for the circular disk of charge of
radius a, but in the present case, assume the surface charge
density to vary with r as
ρs = ρs0 r 2
(C/m2 )
where ρs0 is a constant.
4.18 Multiple charges at different locations are said to be in
equilibrium if the force acting on any one of them is identical
in magnitude and direction to the force acting on any of the
others. Suppose we have two negative charges: one located
at the origin and carrying charge −9e and the other located
on the positive x axis at a distance d from the first one and
carrying charge −36e. Determine the location, polarity, and
magnitude of a third charge whose placement would bring the
entire system into equilibrium.
Section 4-4: Gauss’s Law
∗ 4.19 Three infinite lines of charge, all parallel to the z axis,
are located at the three corners of the kite-shaped arrangement
4.20 Three infinite lines of charge, ρl1 = 3 (nC/m), ρl2 = −3
(nC/m), and ρl3 = 3 (nC/m) are all parallel to the z axis. If they
pass through the respective points (0, −b), (0, 0), and (0, b) in
the x–y plane, find the electric field at (a, 0, 0). Evaluate your
result for a = 2 cm and b = 1 cm.
4.21 A horizontal strip lying in the x–y plane is of width d
in the y direction and infinitely long in the x direction. If the
strip is in air and has a uniform charge distribution ρs , use
Coulomb’s law to obtain an explicit expression for the electric
field at a point P located at a distance h above the centerline
of the strip. Extend your result to the special case where d is
infinite and compare it with Eq. (4.25).
4.22
Given the electric flux density
D = x̂2(x + y) + ŷ(3x − 2y)
(C/m2 ),
determine
(a) ρv by applying Eq. (4.26),
(b) the total charge Q enclosed in a cube 2 m on a side,
located in the first octant with three of its sides coincident
with the x, y, and z axes and one of its corners at the origin,
and
(c) the total charge Q in the cube, obtained by applying
Eq. (4.29).
PROBLEMS
221
∗ 4.23 Repeat Problem 4.22 for D = x̂xy3 z3 (C/m2 ).
4.24 Charge Q1 is uniformly distributed over a thin spherical
shell of radius a, and charge Q2 is uniformly distributed over
a second spherical shell of radius b with b > a. Apply Gauss’s
law to find E in the regions R < a, a < R < b, and R > b.
∗ 4.25 The electric flux density inside a dielectric sphere of
charge density given by
ρv = −
ρv0
,
R2
a ≤ R ≤ b,
where ρv0 is a positive constant, determine D in all regions.
radius a centered at the origin is given by
(C/m2 )
D = R̂ρ0 R
where ρ0 is a constant. Find the total charge inside the sphere.
4.26 In a certain region of space, the charge density is given
in cylindrical coordinates by the function
ρv = 5re−r
(C/m3 ).
Section 4-5: Electric Potential
∗ 4.30 A square in the x–y plane in free space has a point
charge of +Q at corner (a/2, a/2), the same at corner
(a/2, −a/2), and a point charge of −Q at each of the other
two corners.
(a) Find the electric potential at any point P along the x axis.
(b) Evaluate V at x = a/2.
Apply Gauss’s law to find D.
∗ 4.27 An infinitely long cylindrical shell extending between
r = 1 m and r = 3 m contains a uniform charge density ρv0 .
Apply Gauss’s law to find D in all regions.
4.28 If the charge density increases linearly with distance
from the origin such that ρv = 0 at the origin and ρv = 4 C/m3
at R = 2 m, find the corresponding variation of D.
4.29 A spherical shell with outer radius b surrounds a chargefree cavity of radius a < b (Fig. P4.29). If the shell contains a
4.31 The circular disk of radius a shown in Fig. 4-7 has
uniform charge density ρs across its surface.
(a) Obtain an expression for the electric potential V at a point
P(0, 0, z) on the z axis.
(b) Use your result to find E and then evaluate it for z = h.
Compare your final expression with (4.24), which was
obtained on the basis of Coulomb’s law.
∗ 4.32 A circular ring of charge of radius a lies in the x–y plane
and is centered at the origin. Assume also that the ring is in air
and carries a uniform density ρℓ .
(a) Show that the electrical potential at (0, 0, z) is given by
V = ρℓ a/[2ε0 (a2 + z2 )1/2 ].
(b) Find the corresponding electric field E.
r3
ρv
a
r1
r2
b
Figure P4.29 Problem 4.29.
4.33 Show that the electric potential difference V12 between
two points in air at radial distances r1 and r2 from an
infinite line of charge with density ρℓ along the z axis is
V12 = (ρℓ /2πε0 ) ln(r2 /r1 ).
∗ 4.34 Find the electric potential V at a location a distance b
from the origin in the x–y plane due to a line charge with charge
density ρℓ and of length l. The line charge is coincident with
the z axis and extends from z = −l/2 to z = l/2.
4.35 For the electric dipole shown in Fig. 4-15, d = 1 cm and
|E| = 4 (mV/m) at R = 1 m and θ = 0◦ . Find E at R = 2 m and
θ = 90◦ .
222
CHAPTER 4
4.36 For each of the distributions of the electric potential V
shown in Fig. P4.36, sketch the corresponding distribution of
E (in all cases, the vertical axis is in volts and the horizontal
axis is in meters).
y
P = (x, y)
r''
r'
ρl
−ρl
V
(a)
x
(a, 0)
(−a, 0)
30
ELECTROSTATICS
Figure P4.37 Problem 4.37.
3
5
8
11 13
16
x
4.38 Given the electric field
−30
E = R̂
V
18
R2
(V/m)
find the electric potential of point A with respect to point B
where A is at +2 m and B at −4 m (both on the z axis).
(b)
4
∗ 4.39 An infinitely long line of charge with uniform density
3
6
9
12
15
x
ρl = 9 (nC/m) lies in the x–y plane parallel to the y axis
at x = 2 m. Find the potential VAB at point A(3 m, 0, 4 m)
in Cartesian coordinates with respect to point B(0, 0, 0) by
applying the result of Problem 4.33.
4.40 The x–y plane contains a uniform sheet of charge with
ρs1 = 0.2 (nC/m2 ). A second sheet with ρs2 = −0.2 (nC/m2 )
occupies the plane z = 6 m. Find VAB, VBC , and VAC for
A(0, 0, 6 m), B(0, 0, 0), and C(0, −2 m, 2 m).
−4
V
(c)
4
Section 4-6: Conductors
3
6
9
12
15
x
4.41 A cylindrical bar of silicon has a radius of 4 mm and
a length of 8 cm. If a voltage of 5 V is applied between the
ends of the bar and µe = 0.13 (m2 /V·s), µh = 0.05 (m2 /V·s),
Ne = 1.5 × 1016 electrons/m3, and Nh = Ne , find the following:
(a) the conductivity of silicon,
−4
Figure P4.36 Electric potential distributions of Problem 4.36.
(b) the current I flowing in the bar,
∗ (c) the drift velocities u and u ,
e
h
(d) the resistance of the bar, and
(e) the power dissipated in the bar.
∗ 4.37 As shown in Fig. P4.37, two infinite lines of charge,
both parallel to the z axis, lie in the x–z plane: one with
density ρℓ and located at x = a and the other with density −ρℓ
and located at x = −a. Obtain an expression for the electric
potential V (x, y) at a point P = (x, y) relative to the potential at
the origin.
4.42 Repeat Problem 4.41 for a bar of germanium with
µe = 0.4 (m2 /V·s), µh = 0.2 (m2 /V·s), and Ne = Nh =
2.4 × 1019 electrons or holes/m3.
4.43 A 100-m long conductor of uniform cross-section has
a voltage drop of 4 V between its ends. If the density of the
current flowing through it is 1.4 × 106 (A/m2 ), identify the
material of the conductor.
PROBLEMS
223
tangential component of the electric field in the region r ≥ a is
given by Et = −φ̂φ cos φ /r2 , find E in that region.
4.44 A coaxial resistor of length l consists of two concentric cylinders. The inner cylinder has radius a and is made
of a material with conductivity σ1 , and the outer cylinder,
extending between r = a and r = b, is made of a material with
conductivity σ2 . If the two ends of the resistor are capped with
conducting plates, show that the resistance between the two
ends is R = l/[π (σ1 a2 + σ2 (b2 − a2 ))].
∗ 4.50 If E = R̂150 (V/m) at the surface of a 5 cm conducting
of a 20 cm long hollow cylinder (Fig. P4.45) made of carbon
with σ = 3 × 104 (S/m).
4.51 Figure P4.51 shows three planar dielectric slabs of
equal thickness but with different dielectric constants. If E0
in air makes an angle of 45◦ with respect to the z axis, find the
angle of E in each of the other layers.
∗ 4.45 Apply the result of Problem 4.44 to find the resistance
3 cm
sphere centered at the origin, what is the total charge Q on the
sphere’s surface?
z
2 cm
E0
45°
ε0 (air)
Carbon
ε1 = 3ε0
Figure P4.45 Cross-section of hollow cylinder of
Problem 4.45.
ε2 = 5ε0
4.46 A 2 × 10−3 mm thick square sheet of aluminum has
5 cm × 5 cm faces. Find the following:
(a) The resistance between opposite edges on a square face.
(b) The resistance between the two square faces. (See Appendix B for the electrical constants of materials.)
4.47 A cylinder-shaped carbon resistor is 8 cm in length, and
its circular cross section has a diameter d = 1 mm.
(a) Determine the resistance R.
(b) To reduce its resistance by 40%, the carbon resistor is
coated with a layer of copper of thickness t. Use the result
of Problem 4.44 to determine t.
Section 4-8: Boundary Conditions
∗ 4.48 With reference to Fig. 4-21, find E if
1
E2 = x̂3 − ŷ2 + ẑ2
(V/m),
ε1 = 2ε0 , ε2 = 18ε0 , and the boundary has a surface charge
density ρs = 3.54 × 10−11 (C/m2 ). What angle does E2 make
with the z axis?
4.49 An infinitely long cylinder of radius a is surrounded
by a dielectric medium that contains no free charges. If the
ε 3 = 7ε 0
ε0 (air)
Figure P4.51 Dielectric slabs in Problem 4.51.
Sections 4-9 and 4-10: Capacitance and Electrical Energy
4.52 Determine the force of attraction in a parallel-plate
capacitor with A = 5 cm2 , d = 2 cm, and εr = 4 if the voltage
across it is 50 V.
4.53 Dielectric breakdown occurs in a material whenever the magnitude of the field E exceeds the dielectric
strength anywhere in that material. In the coaxial capacitor of
Example 4-15,
∗ (a) At what value of r is |E| maximum?
(b) What is the breakdown voltage if a = 1 cm, b = 2 cm,
and the dielectric material is mica with εr = 6?
224
CHAPTER 4
4.54 An electron with charge Qe = −1.6 × 10−19 C and mass
me = 9.1 × 10−31 kg is injected at a point adjacent to the
negatively charged plate in the region between the plates of
an air-filled, parallel-plate capacitor with separation of 1 cm
and rectangular plates each 10 cm2 in area (Fig. P4.54). If the
voltage across the capacitor is 10 V, find the following:
(a) the force acting on the electron,
(b) the acceleration of the electron, and
(c) the time it takes the electron to reach the positively
charged plate, assuming that it starts from rest.
A1
d
ε1
ELECTROSTATICS
A2
+
V
−
ε2
(a)
1 cm
+
C1
Qe
V
C2
−
(b)
–
+
V0 = 10 V
Figure P4.56 (a) Capacitor with parallel dielectric section and
(b) equivalent circuit.
Figure P4.54 Electron between charged plates of
Problem 4.54.
∗ 4.55 In a dielectric medium with ε = 4, the electric field is
r
given by
E = x̂(x2 + 2z) + ŷx2 − ẑ(y + z)
(V/m)
Calculate the electrostatic energy stored in the region
−1 m ≤ x ≤ 1 m, 0 ≤ y ≤ 2 m, and 0 ≤ z ≤ 3 m.
4.56 Figure P4.56(a) depicts a capacitor consisting of two
parallel, conducting plates separated by a distance d. The space
between the plates contains two adjacent dielectrics: one with
permittivity ε1 and surface area A1 and another with ε2 and A2 .
The objective of this problem is to show that the capacitance
C of the configuration shown in Fig. P4.56(a) is equivalent
to two capacitances in parallel, as illustrated in Fig. P4.56(b),
with
C = C1 + C2 ,
(4.146)
where
ε1 A1
,
d
ε2 A2
.
C2 =
d
C1 =
To this end, proceed as follows:
(a) Find the electric fields E1 and E2 in the two dielectric
layers.
(4.147)
(b) Calculate the energy stored in each section and use the
result to calculate C1 and C2 .
(c) Use the total energy stored in the capacitor to obtain an
expression for C. Show that Eq. (4.146) is indeed a valid
result.
4.57 Use the result of Problem 4.56 to determine the capacitance for each of the following configurations.
∗ (a) Conducting plates are on top and bottom faces of the
rectangular structure in Fig. P4.57(a).
(b) Conducting plates are on front and back faces of the
structure in Fig. P4.57(a).
(4.148)
(c) Conducting plates are on top and bottom faces of the
cylindrical structure in Fig. P4.57(b).
PROBLEMS
225
4.58 The capacitor shown in Fig. P4.58 consists of two
parallel dielectric layers. Use energy considerations to show
that the equivalent capacitance of the overall capacitor, C,
is equal to the series combination of the capacitances of the
individual layers, C1 and C2 , namely
1 cm
3 cm
5 cm
εr = 2
C=
2 cm
εr = 4
(a)
r1 = 2 mm
r2 = 4 mm
r3 = 8 mm
C1C2
,
C1 + C2
(4.149)
where
A
A
,
C2 = ε2
.
d1
d2
(a) Let V1 and V2 be the electric potentials across the upper
and lower dielectrics, respectively. What are the corresponding electric fields E1 and E2 ? By applying the
appropriate boundary condition at the interface between
the two dielectrics, obtain explicit expressions for E1 and
E2 in terms of ε1 , ε2 , and V and the indicated dimensions
of the capacitor.
(b) Calculate the energy stored in each of the dielectric layers
and then use the sum to obtain an expression for C.
(c) Show that C is given by Eq. (4.149).
C1 = ε1
ε3 ε2 ε1
d1
A
d2
+
−
ε1
V
ε2
2 cm
(a)
C1
+
C2
−
V
ε1 = 8ε0; ε2 = 4ε0; ε3 = 2ε0
(b)
(b)
Figure P4.57 Dielectric sections for Problems 4.57 and 4.59.
Figure P4.58 (a) Capacitor with parallel dielectric layers and
(b) equivalent circuit (Problem 4.58).
4.59 Use the expressions given in Problem 4.58 to determine
the capacitance for the configurations in Fig. P4.57(a) when
the conducting plates are placed on the right and left faces of
the structure.
226
CHAPTER 4
4.60 A coaxial capacitor consists of two concentric, conducting, cylindrical surfaces: one of radius a and another
of radius b, as shown in Fig. P4.60. The insulating layer
separating the two conducting surfaces is divided equally into
two semi-cylindrical sections: one filled with dielectric ε1 and
the other filled with dielectric ε2 .
ELECTROSTATICS
z
P = (0, y, z)
Q = (0, d, d)
d
(a) Develop an expression for C in terms of the length l and
the given quantities.
∗ (b) Evaluate the value of C for a = 2 mm, b = 6 mm, ε = 2,
r1
y
d
εr2 = 4, and l = 4 cm.
Figure P4.61 Charge Q next to two perpendicular, grounded,
ε2
a
conducting half-planes.
b
ε1
I2
E
I1
l
(a)
Figure P4.62
(b)
Currents above a conducting plane (Prob-
lem 4.62).
−
V
+
Figure P4.60 Problem 4.60.
∗ 4.63 Use the image method to find the capacitance per unit
length of an infinitely long conducting cylinder of radius a
situated at a distance d from a parallel conducting plane, as
shown in Fig. P4.63.
Section 4-12: Image Method
a
4.61 With reference to Fig. P4.61, charge Q is located
at a distance d above a grounded half-plane located in the
x–y plane and at a distance d from another grounded half-plane
in the x–z plane. Use the image method to
d
(a) Establish the magnitudes, polarities, and locations of the
images of charge Q with respect to each of the two ground
planes (as if each is infinite in extent).
(b) Find the electric potential and electric field at an arbitrary
point P = (0, y, z).
4.62 Conducting wires above a conducting plane carry currents I1 and I2 in the directions shown in Fig. P4.62. Keeping
in mind that the direction of a current is defined in terms of the
movement of positive charges, what are the directions of the
image currents corresponding to I1 and I2 ?
V=0
Figure P4.63 Conducting cylinder above a conducting plane
(Problem 4.63).
Chapter
5
Magnetostatics
Chapter Contents
5-1
5-2
5-3
5-4
TB10
5-5
5-6
5-7
5-8
TB11
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Overview, 228
Magnetic Forces and Torques, 228
The Biot–Savart Law, 236
Maxewell’s Magnetostatic Equations, 242
Vector Magnetic Potential, 246
Electromagnets, 247
Magnetic Properties of Materials, 251
Magnetic Boundary Conditions, 254
Inductance, 256
Magnetic Energy, 261
Inductive Sensors, 262
Chapter 5 Summary, 264
Problems, 265
1. Calculate the magnetic force on a current-carrying wire
placed in a magnetic field and the torque exerted on a
current loop.
2. Apply the Biot–Savart law to calculate the magnetic field
due to current distribution.
3. Apply Ampère’s law to configurations with appropriate
symmetry.
4. Explain magnetic hysteresis in ferromagnetic materials.
5. Calculate the inductance of a solenoid, a coaxial transmission line, or other configurations.
6. Relate the magnetic energy stored in a region to the
magnetic field distribution in that region.
227
Overview
This chapter on magnetostatics parallels the preceding one
on electrostatics. Stationary charges produce static electric
fields, and steady (i.e., non–time-varying) currents produce
static magnetic fields. When ∂ /∂ t = 0, the magnetic fields in
a medium with magnetic permeability µ are governed by the
second pair of Maxwell’s equations (Eqs. (4.3a,b)):
+q
Fm = quB sin θ
θ B
u
(a)
∇ · B = 0,
(5.1a)
× H = J,
∇×
(5.1b)
B
where J is the current density. The magnetic flux density B
and the magnetic field intensity H are related by
B = µ H.
F
(5.2)
−
When examining electric fields in a dielectric medium in
Chapter 4, we noted that the relation D = ε E is valid only when
the medium is linear and isotropic. These properties, which
hold true for most materials, allow us to treat the permittivity ε
as a constant, scalar quantity that is independent of both the
magnitude and the direction of E. A similar statement applies
to the relation given by Eq. (5.2). With the exception of
ferromagnetic materials, for which the relationship between
B and H is nonlinear, most materials are characterized by
constant permeabilities.
u
u
+
F
(b)
Figure 5-1 The direction of the magnetic force exerted on a
charged particle moving in a magnetic field is (a) perpendicular
to both B and u and (b) depends on the charge polarity (positive
or negative).
◮ Furthermore, µ = µ0 for most dielectrics and metals
(excluding ferromagnetic materials). ◭
velocity u through that point. Experiments revealed that a
particle of charge q moving with velocity u in a magnetic field
experiences a magnetic force Fm given by
The objectives of this chapter are to develop an understanding
of the relationship between steady currents and the magnetic
flux B and field H due to various types of current distributions
and in various types of media and to introduce a number of
related quantities, such as the magnetic vector potential A,
the magnetic energy density wm , and the inductance of a conducting structure, L. The parallelism that exists between these
magnetostatic quantities and their electrostatic counterparts is
elucidated in Table 5-1.
×B
Fm = qu×
(N).
(5.3)
Accordingly, the strength of B is measured in newtons/
(C·m/s), which is also called the tesla (T). For a positively
charged particle, the direction of Fm is that of the cross product
u × B, which is perpendicular to the plane containing u and B
and governed by the right-hand rule. If q is negative, the
direction of Fm is reversed (Fig. 5-1). The magnitude of Fm
is given by
(5.4)
Fm = quB sin θ ,
5-1 Magnetic Forces and Torques
where θ is the angle between u and B.
The electric field E at a point in space has been defined as
the electric force Fe per unit charge acting on a charged test
particle placed at that point. We now define the magnetic
flux density B at a point in space in terms of the magnetic
force Fm that acts on a charged test particle moving with
◮ We note that Fm is maximum when u is perpendicular
to B (θ = 90◦ ) and zero when u is parallel to B (θ = 0 or
180◦). ◭
228
5-1 MAGNETIC FORCES AND TORQUES
229
Table 5-1 Attributes of electrostatics and magnetostatics.
Attribute
Electrostatics
Magnetostatics
Sources
Stationary charges ρv
Steady currents J
Fields and Fluxes
E and D
H and B
Constitutive parameter(s)
ε and σ
µ
∇ · D = ρv
×E = 0
∇×
Z
D · ds = Q
∇·B = 0
×H = J
∇×
Z
Z
Z
Governing equations
• Differential form
• Integral form
S
C
Potential
E · dl = 0
Scalar V , with
E = −∇V
S
B · ds = 0
C
H · dl = I
Vector A, with
×A
B = ∇×
Energy density
we = 21 ε E 2
wm = 21 µ H 2
Force on charge q
Fe = qE
×B
Fm = qu×
Circuit element(s)
C and R
If a charged particle resides in the presence of both an
electric field E and a magnetic field B, then the total electromagnetic force acting on it is
× B = q(E + u×
× B).
F = Fe + Fm = qE + qu×
L
Hence, the work performed when a particle is displaced by a
differential distance dl = u dt is
dW = Fm · dl = (Fm · u) dt = 0.
(5.6)
(5.5)
The force expressed by Eq. (5.5) also is known as the Lorentz
force. Electric and magnetic forces exhibit a number of important differences:
1. Whereas the electric force is always in the direction of the
electric field, the magnetic force is always perpendicular
to the magnetic field.
2. Whereas the electric force acts on a charged particle
whether or not it is moving, the magnetic force acts on
it only when it is in motion.
3. Whereas the electric force expends energy in displacing a
charged particle, the magnetic force does no work when
a particle is displaced.
This last statement requires further elaboration. Because the
magnetic force Fm is always perpendicular to u, Fm · u = 0.
◮ Since no work is done, a magnetic field cannot change
the kinetic energy of a charged particle; the magnetic field
can change the direction of motion of a charged particle,
but not its speed. ◭
Exercise 5-1: An electron moving in the positive
x direction perpendicular to a magnetic field is deflected
in the negative z direction. What is the direction of the
magnetic field?
Answer: Positive y direction. (See
EM
.)
Exercise 5-2: A proton moving with a speed of
2 × 106 m/s through a magnetic field with magnetic flux
density of 2.5 T experiences a magnetic force of magnitude 4 × 10−13 N. What is the angle between the magnetic
field and the proton’s velocity?
230
CHAPTER 5
Answer: θ = 30◦ or 150◦ . (See
EM
MAGNETOSTATICS
.)
B
B
Exercise 5-3: A charged particle with velocity u is mov-
ing in a medium with uniform fields E = x̂E and B = ŷB.
What should u be so that the particle experiences no net
force?
u = ẑE/B. (u may also have an arbitrary
y component uy .) (See EM .)
Answer:
I=0
I
(a)
(b)
5-1.1 Magnetic Force on a Current-Carrying
Conductor
A current flowing through a conducting wire consists of
charged particles drifting through the material of the wire.
Consequently, when a current-carrying wire is placed in a
magnetic field, it experiences a force equal to the sum of the
magnetic forces acting on the charged particles moving within
it. Consider, for example, the arrangement shown in Fig. 5-2
in which a vertical wire oriented along the z direction is placed
in a magnetic field B (produced by a magnet) oriented along
the −x̂ direction (into the page). With no current flowing in
the wire, Fm = 0 and the wire maintains its vertical orientation
(Fig. 5-2(a)), but when a current is introduced in the wire, the
wire deflects to the left (−ŷ direction) if the current direction
is upward (+ẑ direction) and to the right (+ŷ direction) if the
current direction is downward (−ẑ direction). The directions
of these deflections are in accordance with the cross product
given by Eq. (5.3).
To quantify the relationship between Fm and the current I
flowing in a wire, let us consider a small segment of the
wire of cross-sectional area A and differential length dl with
the direction of dl denoting the direction of the current.
Without loss of generality, we assume that the charge carriers
constituting the current I are exclusively electrons, which is
always a valid assumption for a good conductor. If the wire
contains a free-electron charge density ρve = −Ne e, where
Ne is the number of moving electrons per unit volume, then
the total amount of moving charge contained in an elemental
volume of the wire is
dQ = ρve A dl = −Ne eA dl,
(5.7)
and the corresponding magnetic force acting on dQ in the
presence of a magnetic field B is
dFm = dQ ue × B = −Ne eA dl ue × B,
(5.8a)
where ue is the drift velocity of the electrons. Since the
direction of a current is defined as the direction of flow of
positive charges, the electron drift velocity ue is parallel to dl,
B
z
x
y
I
(c)
Figure 5-2 When a slightly flexible vertical wire is placed in a
magnetic field directed into the page (as denoted by the crosses),
it is (a) not deflected when the current through it is zero, (b)
deflected to the left when I is upward, and (c) deflected to the
right when I is downward.
but opposite in direction. Thus, dl ue = −dl ue and Eq. (5.8a)
becomes
× B.
dFm = Ne eAue dl×
(5.8b)
From Eqs. (4.11) and (4.12), the current I flowing through a
cross-sectional area A due to electrons with density ρve = −Ne e
and moving with velocity −ue is
I = ρve (−ue )A = (−Ne e)(−ue )A = Ne eAue .
Hence, Eq. (5.8b) may be written in the compact form
×B
dFm = I dl×
(N).
(5.9)
For a closed circuit of contour C carrying a current I, the total
magnetic force is
Fm = I
Z
C
×B
dl×
(N).
(5.10)
5-1 MAGNETIC FORCES AND TORQUES
231
Module 5.1 Electron Motion in Static Fields This module demonstrates the Lorentz force on an electron moving under
the influence of an electric field alone, a magnetic field alone, or both acting simultaneously.
If the closed wire shown in Fig. 5-3(a) resides in a uniform
external magnetic field B, then B can be taken outside the
integral in Eq. (5.10), in which case
Z dl × B = 0.
(5.11)
Fm = I
C
◮ This result, which is a consequence of the fact that the
vector sum of the infinitesimal vectors dl over a closed
path equals zero, states that the total magnetic force on
any closed current loop in a uniform magnetic field is
zero. ◭
In the study of magnetostatics, all currents flow through
closed paths. To understand why, consider the curved wire in
Fig. 5-3(b) carrying a current I from point a to point b. In doing
so, negative charges accumulate at a, and positive ones at b.
The time-varying nature of these charges violates the static
assumptions underlying Eqs. (5.1a,b).
If we are interested in the magnetic force exerted on a wire
segment l (Fig. 5-3(b)) residing in a uniform magnetic field
(while realizing that it is part of a closed current loop), we can
integrate Eq. (5.9) to obtain
Z Fm = I
dl × B = Iℓℓ × B,
(5.12)
ℓ
This does not mean that the force at every point along the wire
is zero, but rather that the vector sum of all forces exerted on
all parts of the closed wire adds up to zero.
where ℓ is the vector directed from a to b (Fig. 5-3(b)). The
integral of dl from a to b has the same value irrespective of the
path taken between a and b. For a closed loop, points a and b
232
CHAPTER 5
MAGNETOSTATICS
y
C
B
φ
I
B
dl
dφ r
φ
x
I
dl
Figure 5-4 Semicircular conductor in a uniform field (Example 5-1).
(a)
x–y plane and carries a current I. The closed circuit is exposed
to a uniform magnetic field B = ŷB0 . Determine (a) the
magnetic force F1 on the straight section of the wire and (b)
the force F2 on the curved section.
b
B
I
Vector l
Solution: (a) To evaluate F1 , consider that the straight section
of the circuit is of length 2r and its current flows along the +x
direction. Application of Eq. (5.12) with ℓ = x̂ 2r gives
× ŷB0 = ẑ 2IrB0
F1 = x̂(2Ir)×
dl
a
(b)
Figure 5-3 In a uniform magnetic field, (a) the net force
on a closed current loop is zero because the integral of the
displacement vector dl over a closed contour is zero and (b) the
force on a line segment is proportional to the vector between the
end point (Fm = Iℓℓ × B).
The ẑ direction in Fig. 5-4 is out of the page.
(b) To evaluate F2 , consider a segment of differential length dl
on the curved part of the circle. The direction of dl is chosen
to coincide with the direction of the current. Since dl and B
are both in the x–y plane, their cross product dl × B points
in the negative z direction, and the magnitude of dl × B is
proportional to sin φ , where φ is the angle between dl and B.
Moreover, the magnitude of dl is dl = r d φ . Hence,
F2 = I
Z π
φ =0
= −ẑI
become the same point, in which case ℓ = 0 and Fm = 0.
Example 5-1:
Force on a Semicircular
Conductor
×B
dl×
Z π
φ =0
rB0 sin φ d φ = −ẑ2IrB0
(N).
The −ẑ direction of the force acting on the curved part of the
conductor is into the page. We note that F2 = −F1 , implying
that no net force acts on the closed loop, although opposing
forces act on its two sections.
What are the major differences
between the electric force Fe and the magnetic force Fm ?
Concept Question 5-1:
The semicircular conductor shown in Fig. 5-4 lies in the
(N).
5-1 MAGNETIC FORCES AND TORQUES
233
The ends of a 10-cm long wire
carrying a constant current I are anchored at two points
on the x axis, namely x = 0 and x = 6 cm. If the wire lies
in the x–y plane in a magnetic field B = ŷB0 , which of
the following arrangements produces a greater magnetic
force on the wire: (a) wire is V-shaped with corners at
(0, 0), (3, 4), and (6, 0) or (b) wire is an open rectangle
with corners at (0, 0), (0, 2), (6, 2), and (6, 0)?
Concept Question 5-2:
Exercise 5-4: A horizontal wire with a mass per unit
length of 0.2 kg/m carries a current of 4 A in the
+x direction. If the wire is placed in a uniform magnetic
flux density B, what should the direction and minimum
magnitude of B be in order to magnetically lift the wire
vertically upward? (Hint: The acceleration due to gravity
is g = −ẑ9.8 m/s2 .)
Answer: B = ŷ0.49 T. (See
EM
.)
5-1.2 Magnetic Torque on a Current-Carrying
Loop
When a force is applied on a rigid body that can pivot about a
fixed axis, the body will, in general, react by rotating about that
axis. The angular acceleration depends on the cross product of
the applied force vector F and the distance vector d, which
is measured from a point on the rotation axis (such that d
is perpendicular to the axis) to the point of application of F
(Fig. 5-5). The length of d is called the moment arm, and the
cross product
×F
T = d×
(N·m)
is called the torque. The unit for T is the same as that for work
or energy, even though torque does not represent either. The
force F applied on the disk shown in Fig. 5-5 lies in the x–y
plane and makes an angle θ with d. Hence,
T = ẑrF sin θ ,
(5.14)
where |d| = r, which is the radius of the disk, and F = |F|.
From Eq. (5.14), we observe that a torque along the positive
z direction corresponds to a tendency for the cylinder to
rotate counterclockwise and, conversely, a torque along the
−z direction corresponds to clockwise rotation.
◮ These directions are governed by the following righthand rule: When the thumb of the right hand points along
the direction of the torque, the four fingers indicate the
direction that the torque tries to rotate the body. ◭
We now consider the magnetic torque exerted on a conducting loop under the influence of magnetic forces. We begin with
the simple case where the magnetic field B is in the plane of
the loop, and then we extend the analysis to the more general
case where B makes an angle θ with the surface normal of the
loop.
Magnetic Field in the Plane of the Loop
The rectangular conducting loop shown in Fig. 5-6(a) is
constructed from rigid wire and carries a current I. The loop
lies in the x–y plane and is allowed to pivot about the axis
shown. Under the influence of an externally generated uniform
magnetic field B = x̂B0 , arms 1 and 3 of the loop are subjected
to forces F1 and F3 given by
(5.13)
× (x̂B0 ) = ẑIbB0 ,
F1 = I(−ŷb)×
(5.15a)
and
y
x
Pivot axis
d
θ
T
z
F
Figure 5-5 The force F acting on a circular disk that can pivot
along the z axis generates a torque T = d× F that causes the disk
to rotate.
× (x̂B0 ) = −ẑIbB0 .
F3 = I(ŷb)×
(5.15b)
These results are based on the application of Eq. (5.12). We
note that the magnetic forces acting on arms 1 and 3 are in
opposite directions, and no magnetic force is exerted on either
arm 2 or 4 because B is parallel to the direction of the current
flowing in those arms.
A bottom view of the loop, depicted in Fig. 5-6(b), reveals
that forces F1 and F3 produce a torque about the origin O,
causing the loop to rotate in a clockwise direction. The moment arm is a/2 for both forces, but d1 and d3 are in opposite
directions, resulting in a total magnetic torque of
T = d1 × F1 + d3 × F3
a
a
× (ẑIbB0 ) + x̂
× (−ẑIbB0 )
= −x̂
2
2
= ŷIabB0 = ŷIAB0 ,
(5.16)
234
CHAPTER 5
MAGNETOSTATICS
y
z
B
2
I
Pivot axis
y
B
F1
O
1
3
F2
x
b
nˆ
1
B
4
θ
a
4
Pivot axis
3
I
(a)
a
z
F1
b
F4
z
a/2
y
Loop arm 1
O
d1
2
d3
x
(a)
x
m (magnetic
moment)
F1
Loop arm 3
nˆ
F3
B
B
F3
−z
(b)
Figure 5-6 Rectangular loop pivoted along the y axis: (a) front
Arm 1
θ
a/2
θ
B
(a/2) sin θ O
Arm 3
view and (b) bottom view. The combination of forces F1 and F3
on the loop generates a torque that tends to rotate the loop in a
clockwise direction as shown in (b).
F3
(b)
where A = ab is the area of the loop. The right-hand rule tells
us that the sense of rotation is clockwise. The result given by
Eq. (5.16) is valid only when the magnetic field B is parallel to
the plane of the loop. As soon as the loop starts to rotate, the
torque T decreases, and at the end of one quarter of a complete
rotation, the torque becomes zero, as discussed next.
Magnetic Field Perpendicular to the Axis of a Rectangular
Loop
For the situation represented by Fig. 5-7, where B = x̂B0 , the
field is still perpendicular to the loop’s axis of rotation, but
because its direction may be at any angle θ with respect to
the loop’s surface normal n̂, we may now have nonzero forces
on all four arms of the rectangular loop. However, forces F2
Figure 5-7 Rectangular loop in a uniform magnetic field with
flux density B whose direction is perpendicular to the rotation
axis of the loop but makes an angle θ with the loop’s surface
normal n̂.
and F4 are equal in magnitude and opposite in direction and
are along the rotation axis; hence, the net torque contributed
by their combination is zero. The directions of the currents in
arms 1 and 3 are always perpendicular to B regardless of the
magnitude of θ . Hence, F1 and F3 have the same expressions
given previously by Eqs. (5.15a,b), and for 0 ≤ θ ≤ π /2, their
moment arms are of magnitude (a/2) sin θ , as illustrated in
Fig. 5-7(b). Consequently, the magnitude of the net torque
5-1 MAGNETIC FORCES AND TORQUES
235
Module 5.2 Magnetic Fields Due to Line Sources You can place z-directed linear currents anywhere in the display
plane (x–y plane), select their magnitudes and directions, and then observe the spatial pattern of the induced magnetic flux
B(x, y).
exerted by the magnetic field about the axis of rotation is the
same as that given by Eq. (5.16), but modified by sin θ :
T = IAB0 sin θ .
(5.17)
According to Eq. (5.17), the torque is maximum when the
magnetic field is parallel to the plane of the loop (θ = 90◦ )
and zero when the field is perpendicular to the plane of the loop
(θ = 0). If the loop consists of N turns with each contributing
a torque given by Eq. (5.17), then the total torque is
T = NIAB0 sin θ .
(5.18)
The quantity NIA is called the magnetic moment m of the loop.
Now, consider the vector
where n̂ is the surface normal of the loop and governed by the
following right-hand rule: When the four fingers of the right
hand advance in the direction of the current I, the direction of
the thumb specifies the direction of n̂. In terms of m, the torque
vector T can be written as
×B
T = m×
(A·m2 ),
(5.19)
(5.20)
Even though the derivation leading to Eq. (5.20) was obtained
for B perpendicular to the axis of rotation of a rectangular loop,
the expression is valid for any orientation of B and for a loop
of any shape.
How is the direction of the
magnetic moment of a loop defined?
Concept Question 5-3:
m = n̂ NIA = n̂ m
(N·m).
236
CHAPTER 5
If one of two wires of equal
length is formed into a closed square loop and the other
into a closed circular loop, and if both wires are carrying
equal currents and both loops have their planes parallel
to a uniform magnetic field, which loop would experience
the greater torque?
Concept Question 5-4:
MAGNETOSTATICS
(dH out of the page)
P
dH
R
ˆ
R
I
θ
Exercise 5-5: A square coil of 100 turns and 0.5-m long
dl
sides is in a region with a uniform magnetic flux density
of 0.2 T. If the maximum magnetic torque exerted on the
coil is 4 × 10−2 (N·m), what is the current flowing in the
coil?
Answer: I = 8 mA. (See
EM
P' dH
(dH into the page)
.)
Figure 5-8 Magnetic field dH generated by a current element
I dl. The direction of the field induced at point P is opposite to
that induced at point P′ .
5-2 The Biot–Savart Law
In the preceding section, we elected to use the magnetic flux
density B to denote the presence of a magnetic field in a given
region of space. We now work with the magnetic field intensity
H instead. We do this in part to remind the reader that for most
materials the flux and field are linearly related by B = µ H;
therefore, knowledge of one implies knowledge of the other
(assuming that µ is known).
Through his experiments on the deflection of compass
needles by current-carrying wires, Hans Oersted established
that currents induce magnetic fields that form closed loops
around the wires (see Section 1-3.3). Building upon Oersted’s
results, Jean Biot and Félix Savart arrived at an expression
that relates the magnetic field H at any point in space to the
current I that generates H. The Biot–Savart law states that the
differential magnetic field dH generated by a steady current I
flowing through a differential length vector dl is
dH =
× R̂
I dl×
4 π R2
(A/m),
the direction of dH is out of the page, whereas at point P′ the
direction of dH is into the page.
To determine the total magnetic field H due to a conductor
of finite size, we need to sum up the contributions due to all the
current elements making up the conductor. Hence, the Biot–
Savart law becomes
H=
I
4π
Z
l
× R̂
dl×
R2
(A/m),
(5.22)
where l is the line path along which I exists.
(5.21)
where R = R̂R is the distance vector between dl and the
observation point P, as shown in Fig. 5-8. The SI unit for H
is ampere·m/m2 = (A/m). It is important to remember that
Eq. (5.21) assumes that dl is along the direction of the current I
and the unit vector R̂ points from the current element to the
observation point.
According to Eq. (5.21), dH varies as R−2 , which is similar
to the distance dependence of the electric field induced by an
electric charge. However, unlike the electric field vector E,
whose direction is along the distance vector R joining the
charge to the observation point, the magnetic field H is
orthogonal to the plane containing the direction of the current
element dl and the distance vector R. At point P in Fig. 5-8,
5-2.1
Magnetic Field Due to Surface and Volume
Current Distributions
The Biot–Savart law also may be expressed in terms of
distributed current sources (Fig. 5-9) such as the volume
current density J, measured in (A/m2 ), or the surface current
density Js , measured in (A/m). The surface current density Js
applies to currents that flow on the surface of a conductor in
the form of a sheet of effectively zero thickness. When current
sources are specified in terms of Js over a surface S or in terms
of J over a volume υ , we can use the equivalence given by
I dl
Js ds
J dυ
(5.23)
5-2 THE BIOT–SAVART LAW
237
z
J
I
l
(a)
dl
R
dθ
A/m2
ˆ
R
P
θ
S
(a) Volume current density J in
dl
r
dH into
the page
z
Js
θ2
l
(b)
R2
I
l
r
(b) Surface current density Js in A/m
P
θ1
Figure 5-9 (a) The
R total current crossing the cross section S of
the cylinder is I = S J · ds. (b) TheR total current flowing across
the surface of the conductor is I = l Js dl.
1
4π
Z
Js × R̂
ds,
2
S R
(5.24a)
(surface current)
H=
1
4π
Z
υ
J × R̂
dυ .
R2
Figure 5-10 Linear conductor of length l carrying a current I.
(a) The field dH at point P due to incremental current element
dl. (b) Limiting angles θ1 and θ2 , each measured between
vector I dl and the vector connecting the end of the conductor
associated with that angle to point P (Example 5-2).
to express the Biot–Savart law as
H=
(5.24b)
(volume current)
Solution: From Fig. 5-10, the direction of I is along +ẑ.
Hence, the differential length vector is dl = ẑ dz, and
dl × R̂ = dz (ẑ × R̂) = φ̂φ sin θ̂θ dz, where φ̂φ is the azimuth
direction and θ is the angle between dl and R̂. Application
of Eq. (5.22) gives
H=
Example 5-2:
R1
Magnetic Field of a Linear
Conductor
A free-standing linear conductor of length l carries a current I
along the z axis as shown in Fig. 5-10. Determine the magnetic
flux density B at a point P located at a distance r in the
x–y plane. The wire is, of course, part of a closed-loop circuit,
but our current interest is only in the conducting wire.
I
4π
Z z=l/2
× R̂
dl×
z=−l/2
R2
φ
= φ̂
I
4π
Z l/2
sin θ
−l/2
R2
dz.
(5.25)
Both R and θ are dependent on the integration variable z, but
the radial distance r is not. For convenience, we convert the
integration variable from z to θ by using the transformations
R = r csc θ ,
(5.26a)
z = −r cot θ ,
(5.26b)
dz = r csc θ d θ .
(5.26c)
2
238
CHAPTER 5
Upon inserting Eqs. (5.26a) and (5.26c) into Eq. (5.25), we
have
H = φ̂φ
= φ̂φ
I
Z θ2
I
sin θ r csc2 θ d θ
4π
Z θ2
I
4π r
θ1
Magnetic field
r2 csc2 θ
θ1
sin θ d θ = φ̂φ
I
(cos θ1 − cos θ2 ),
4π r
B
B
(5.27)
B
where θ1 and θ2 are the limiting angles at z = −l/2 and
z = l/2, respectively. From the right triangle in Fig. 5-10(b), it
follows that
l/2
,
cos θ1 = p
r2 + (l/2)2
−l/2
cos θ2 = − cos θ1 = p
.
2
r + (l/2)2
(5.28a)
I
Figure 5-11 Magnetic field surrounding a long, linear currentcarrying conductor.
(5.28b)
z
Hence,
dH'
µ Il
√0
(T).
(5.29)
2π r 4r2 + l 2
For an infinitely long wire with l ≫ r, Eq. (5.29) reduces to
dH'z
dHz
B = µ0 H = φ̂φ
B = φ̂φ
µ0 I
.
2π r
(infinitely long wire)
MAGNETOSTATICS
dH'r
dH
θ
P = (0, 0, z)
(5.30)
dHr
R
◮ This is a very important and useful expression to keep
in mind. It states that in the neighborhood of a linear
conductor carrying a current I, the induced magnetic field
forms concentric circles around the wire (Fig. 5-11), and
its intensity is directly proportional to I and inversely
proportional to the distance r. ◭
dl'
φ+π
y
φ
a
θ
dl
Example 5-3:
Magnetic Field of a Circular Loop
A circular loop of radius a carries a steady current I. Determine
the magnetic field H at a point on the axis of the loop.
Solution: Let us place the loop in the x–y plane (Fig. 5-12).
Our task is to obtain an expression for H at point P(0, 0, z).
We start by noting that any element dl on the circular loop
is perpendicular to the distance vector R and that all elements
around
√ the loop are at the same distance R from P with
R = a2 + z2 . From Eq. (5.21), the magnitude of dH due to
current element dl is
dH =
I dl
I
× R̂| =
|dl×
,
4 π R2
4π (a2 + z2 )
(5.31)
x
I
Figure 5-12 Circular loop carrying a current I (Example 5-3).
and the direction of dH is perpendicular to the plane containing
R and dl. dH is in the r–z plane (Fig. 5-12); therefore, it has
components dHr and dHz . If we consider element dl′ located
diametrically opposite to dl, we observe that the z components
of the magnetic fields due to dl and dl′ add because they are
in the same direction, but their r components cancel because
they are in opposite directions. Hence, the non-vanishing
5-2 THE BIOT–SAVART LAW
239
Module 5.3 Magnetic Field of a Current Loop Examine how the field along the loop axis changes with loop parameters.
At the center of the loop (z = 0), Eq. (5.34) reduces to
component of the magnetic field is along z only. That is,
dH = ẑ dHz = ẑ dH cos θ
H = ẑ
I cos θ
= ẑ
dl.
4π (a2 + z2 )
(5.32)
For a fixed point P(0, 0, z) on the axis of the loop, all quantities
in Eq. (5.32) are constant, except for dl. Hence, integrating
Eq. (5.32) over a circle of radius a gives
H = ẑ
I cos θ
4π (a2 + z2 )
Z
(5.33)
Upon using the relation cos θ = a/(a2 + z2 )1/2 , we obtain
Ia2
2(a2 + z2 )3/2
(A/m).
(5.35)
and at points very far away from the loop such that z2 ≫ a2 ,
Eq. (5.34) simplifies to
H = ẑ
Ia2
2|z|3
(at |z| ≫ a).
(5.36)
In view of the definition given by Eq. (5.19) for the magnetic
moment m of a current loop, a single-turn loop situated in
the x–y plane (Fig. 5-12) has magnetic moment m = ẑm with
m = I π a2. Consequently, Eq. (5.36) may be expressed as
H = ẑ
H = ẑ
(at z = 0),
5-2.2 Magnetic Field of a Magnetic Dipole
dl
I cos θ
(2π a).
= ẑ
4π (a2 + z2 )
I
2a
(5.34)
m
2π |z|3
(at |z| ≫ a).
(5.37)
This expression applies to a point P far away from the loop
and on its axis. Had we solved for H at any distant point
240
CHAPTER 5
P = (R, θ , φ ) in a spherical coordinate system with R the
distance between the center of the loop and point P, we would
have obtained the expression
H=
m
(R̂ 2 cos θ + θ̂θ sin θ )
4 π R3
(for R ≫ a).
MAGNETOSTATICS
Exercise 5-6: A semi-infinite linear conductor extends
between z = 0 and z = ∞ along the z axis. If the current I
in the conductor flows along the positive z direction, find
H at a point in the x–y plane at a radial distance r from the
conductor.
I
φ
(A/m). (See EM .)
Answer: H = φ̂
4π r
(5.38)
Exercise 5-7: A wire carrying a current of 4 A is formed
◮ A current loop with dimensions much smaller than the
distance between the loop and the observation point is
called a magnetic dipole. This is because the pattern of
its magnetic field lines is similar to that of a permanent
magnet as well as to the pattern of the electric field lines
of the electric dipole (Fig. 5-13). ◭
into a circular loop. If the magnetic field at the center of
the loop is 20 A/m, what is the radius of the loop if the
loop has (a) only one turn and (b) 10 turns?
Answer: (a) a = 10 cm, (b) a = 1 m. (See
EM
.)
Exercise 5-8: A wire is formed into a square loop and
placed in the x–y plane with its center at the origin and
each of its sides parallel to either the x or y axes. Each
side is 40 cm in length, and the wire carries a current of
5 A whose direction is clockwise when the loop is viewed
from above. Calculate the magnetic field at the center of
the loop.
Two infinitely long parallel
wires carry currents of equal magnitude. What is the
resultant magnetic field due to the two wires at a point
midway between the wires—compared with the magnetic
field due to one of them alone—if the currents are (a) in
the same direction and (b) in opposite directions?
Concept Question 5-5:
4I
= −ẑ11.25 A/m. (See
2π l
Answer: H = −ẑ √
Devise a right-hand rule for the
direction of the magnetic field due to a linear currentcarrying conductor.
Concept Question 5-6:
EM
.)
What is a magnetic dipole?
Describe its magnetic field distribution.
Concept Question 5-7:
E
H
H
N
+
I
−
S
(a) Electric dipole
(b) Magnetic dipole
(c) Bar magnet
Figure 5-13 Patterns of (a) the electric field of an electric dipole, (b) the magnetic field of a magnetic dipole, and (c) the magnetic field
of a bar magnet. Far away from the sources, the field patterns are similar in all three cases.
5-2 THE BIOT–SAVART LAW
241
Module 5.4 Magnetic Force between Two Parallel Conductors Observe the direction and magnitude of the force
exerted on parallel current-carrying wires.
5-2.3 Magnetic Force Between Two Parallel
Conductors
The force F2 exerted on a length l of wire I2 due to its presence
in field B1 may be obtained by applying Eq. (5.12):
In Section 5-1.1, we examined the magnetic force Fm that
acts on a current-carrying conductor when placed in an external magnetic field. The current in the conductor, however,
also generates its own magnetic field. Hence, if two currentcarrying conductors are placed in each other’s vicinity, each
will exert a magnetic force on the other. Let us consider
two very long (or effectively infinitely long), straight, freestanding, parallel wires separated by a distance d and carrying
currents I1 and I2 in the z direction (Fig. 5-14) at y = −d/2 and
y = d/2, respectively. We denote by B1 the magnetic field due
to current I1 , defined at the location of the wire carrying current
I2 and, conversely, by B2 the field due to I2 at the location of the
wire carrying current I1 . From Eq. (5.30), with I = I1 , r = d,
and φ̂φ = −x̂ at the location of I2 , the field B1 is
B1 = −x̂
µ0 I1
.
2π d
(5.39)
× B1 = I2 l ẑ×
× (−x̂)
F2 = I2 l ẑ×
µ0 I1
µ0 I1 I2 l
= −ŷ
,
2π d
2π d
(5.40)
and the corresponding force per unit length is
F′2 =
F2
µ0 I1 I2
= −ŷ
.
l
2π d
(5.41)
A similar analysis performed for the force per unit length
exerted on the wire carrying I1 leads to
F′1 = ŷ
µ0 I1 I2
.
2π d
(5.42)
◮ Thus, two parallel wires carrying currents in the same
direction attract each other with equal force. If the currents
are in opposite directions, the wires would repel one
another with equal force. ◭
242
CHAPTER 5
that is
R enclosed by a surface S and contains a total charge
Q = υ ρv d υ (Section 4-4).
The magnetostatic counterpart of Eq. (5.43), often called
Gauss’s law for magnetism, is
z
I1
I2
F'1
F'2
d/2
l
d/2
d
∇·B = 0
x
Figure 5-14 Magnetic forces on parallel current-carrying
conductors.
Exercise 5-9: Suppose that the conductor carrying I2 in
Fig. 5-14 is rotated so that it is parallel to the x axis. What
would F2 be in that case?
EM
.)
5-3 Maxwell’s Magnetostatic Equations
Thus far, we have introduced the Biot–Savart law for finding
the magnetic flux density B and field H due to any distribution
of electric currents in free space, and we examined how
magnetic fields can exert magnetic forces on moving charged
particles and current-carrying conductors. We now examine
two additional important properties of magnetostatic fields.
5-3.1 Gauss’s Law for Magnetism
In Chapter 4, we learned that the net outward flux of the electric flux density D through a closed surface equals the enclosed
net charge Q. We referred to this property as Gauss’s law (for
electricity) and expressed it mathematically in differential and
integral forms as
∇ · D = ρv
Z
S
D · ds = Q.
Z
S
B · ds = 0.
(5.44)
The differential form is one of Maxwell’s four equations, and
the integral form is obtained with the help of the divergence
theorem. Note that the right-hand side of Gauss’s law for magnetism is zero, reflecting the fact that the magnetic equivalence
of an electric point charge does not exist in nature.
y
B1
Answer: F2 = 0. (See
MAGNETOSTATICS
(5.43)
Conversion from differential to integral form was accomplished by applying the divergence theorem to a volume υ
◮ The hypothetical magnetic analogue to an electric
point charge is called a magnetic monopole. Magnetic
monopoles, however, always occur in pairs (that is, as
dipoles). ◭
No matter how many times a permanent magnet is subdivided,
each new piece will always have a north and a south pole, even
if the process were to be continued down to the atomic level.
Consequently, there is no magnetic equivalence to an electric
charge q or charge density ρv .
Formally, the name “Gauss’s law” refers to the electric case,
even when no specific reference to electricity is indicated. The
property described by Eq. (5.44) has been called “the law of
nonexistence of isolated monopoles,” “the law of conservation
of magnetic flux,” and “Gauss’s law for magnetism,” among
others. We prefer the last of the three cited names because
it reminds us of the parallelism, as well as the differences,
between the electric and magnetic laws of nature.
The difference between Gauss’s law for electricity and its
magnetic counterpart can be elucidated in terms of field lines.
Electric field lines originate from positive electric charges and
terminate on negative ones. Hence, for the electric field lines
of the electric dipole shown in Fig. 5-15(a), the electric flux
through a closed surface surrounding one of the charges is
nonzero. In contrast, magnetic field lines always form continuous closed loops. As we saw in Section 5-2, the magnetic
field lines due to currents do not begin or end at any point; this
is true for the linear conductor of Fig. 5-11 and the circular
loop of Fig. 5-12, as well as for any current distribution. It is
also true for a bar magnet (Fig. 5-15(b)). Because the magnetic
field lines form closed loops, the net magnetic flux through any
closed surface surrounding the south pole of the magnet (or
through any other closed surface) is always zero, regardless of
its shape.
5-3 MAXWELL’S MAGNETOSTATIC EQUATIONS
243
H
H
H
E
I
I
N
+
C
H
C
H
(a)
−
S
(b)
H
Closed imaginary
surface
H
I
(b) Bar magnet
(a) Electric dipole
Figure 5-15 Whereas (a) the net electric flux through a closed
surface surrounding a charge is not zero, and (b) the net magnetic
flux through a closed surface surrounding one of the poles of a
magnet is zero.
C
(c)
Figure 5-16 Ampère’s law states that the line integral of H
5-3.2 Ampère’s Law
In Chapter 4 we learned that the electrostatic field is conservative, meaning that its line integral along a closed contour
always vanishes. This property of the electrostatic field was
expressed in differential and integral forms as
×E = 0
∇×
Z
C
E · dℓℓ = 0.
(5.45)
Conversion of the differential to integral form was accomplished by applying Stokes’s theorem to a surface S with
contour C.
The magnetostatic counterpart of Eq. (5.45), known as
Ampère’s law, is
×H = J
∇×
Z
C
H · dℓℓ = I,
(5.46)
where I is the total current passing through S. The differential
form again is one of Maxwell’s equations, and the integral
form is obtained by integrating both sides of Eq. (5.46) over
an open surface S,
Z
S
× H) · ds =
(∇×
Z
S
J · ds,
(5.47)
and then invoking Stokes’s theorem with I = J · ds.
R
around a closed contour C is equal to the current traversing the
surface bounded by the contour. This is true for contours (a) and
(b), but the line integral of H is zero for the contour in (c)
because the current I (denoted by the symbol ⊙) is not enclosed
by the contour C.
◮ The sign convention for the direction of the contour
path C in Ampère’s law is taken so that I and H satisfy
the right-hand rule defined earlier in connection with the
Biot–Savart law. That is, if the direction of I is aligned
with the direction of the thumb of the right hand, then the
direction of the contour C should be chosen along that of
the other four fingers. ◭
In words, Ampère’s circuital law states that the line integral
of H around a closed path is equal to the current traversing
the surface bounded by that path. To apply Ampère’s law, the
current must flow through a closed path. By way of illustration,
for both configurations shown in Figs. 5-16(a) and (b), the
line integral of H is equal to the current I, even though the
paths have very different shapes and the magnitude of H is not
uniform along the path of the configuration in part (b). By the
same token, because path (c) in Fig. 5-16 does not enclose the
current I, the line integral of H along it vanishes, even though
H is not zero along the path.
When we examined Gauss’s law in Section 4-4, we discovered that in practice its usefulness for calculating the electric
flux density D is limited to charge distributions that possess a
certain degree of symmetry and that the calculation procedure
244
CHAPTER 5
is subject to the proper choice of a Gaussian surface enclosing
the charges. A similar restriction applies to Ampère’s law: its
usefulness is limited to symmetric current distributions that
allow the choice of convenient Ampèrian contours around
them, as illustrated by Examples 5-4 to 5-6.
MAGNETOSTATICS
Contour C2
for r2 ≥ a
I
r2
C2
r1
C1
Example 5-4:
Magnetic Field of a Long Wire
a
A long (practically infinite) straight wire of radius a carries
a steady current I that is uniformly distributed over its cross
section. Determine the magnetic field H a distance r from the
wire axis for (a) r ≤ a (inside the wire) and (b) r ≥ a (outside
the wire).
z
(a) Cylindrical wire
Solution: (a) We choose I to be along the +z direction
(Fig. 5-17(a)). To determine H1 = H at a distance r = r1 ≤ a,
we choose the Ampèrian contour C1 to be a circular path of
radius r = r1 (Fig. 5-17(b)). In this case, Ampère’s law takes
the form
Z
H1 · dl1 = I1 ,
(5.48)
y
r2
φ̂
C2
C1
C1
H1 · dl1 =
Z 2π
0
(b) For r = r2 ≥ a, we choose path C2 , which encloses all the
current I. Hence, H2 = φ̂φ H2 , dℓℓ2 = φ̂φ r2 d φ , and
H2 · dl2 = 2π r2H2 = I,
which yields
H2 = φ̂φH2 = φ̂φ
I
2π r2
(for r2 ≥ a).
x
(b) Wire cross section
H(r)
Equating both sides of Eq. (5.48) and then solving for H1
yields
r1
φ
H1 = φ̂φH1 = φ̂
I
(for r1 ≤ a).
(5.49a)
2π a2
C2
φ1
a
H1 (φ̂φ · φ̂φ)r1 d φ = 2π r1 H1 .
The current I1 flowing through the area enclosed by C1 is equal
to the total current I multiplied by the ratio of the area enclosed
by C1 to the total cross-sectional area of the wire:
2
r 2
π r1
1
I
=
I1 =
I.
π a2
a
Z
r1
C1
where I1 is the fraction of the total current I flowing
through C1 . From symmetry, H1 must be constant in magnitude and parallel to the contour at any point along the path.
Furthermore, to satisfy the right-hand rule and given that I is
along the z direction, H1 must be in the +φ direction. Hence,
H1 = φ̂φH1 , dl1 = φ̂φr1 d φ , and the left-hand side of Eq. (5.48)
becomes
Z
Contour C1
for r1 ≤ a
(5.49b)
H(a) =
I
2πa
H1
H2
r
a
(c)
Figure 5-17 Infinitely long wire of radius a carrying a uniform
current I along the +z direction: (a) general configuration
showing contours C1 and C2 ; (b) cross-sectional view; and (c)
a plot of H versus r (Example 5-4).
Ignoring the subscript 2, we observe that Eq. (5.49b) provides
the same expression for B = µ0 H as Eq. (5.30), which was
derived on the basis of the Biot–Savart law.
The variation of the magnitude of H as a function of r is
5-3 MAXWELL’S MAGNETOSTATIC EQUATIONS
245
plotted in Fig. 5-17(c); H increases linearly between r = 0 and
r = a (inside the conductor), and then decreases as 1/r for
r > a (outside the conductor).
H
φ̂
Exercise 5-10: A current I flows in the inner conductor
of a long coaxial cable and returns through the outer
conductor. What is the magnetic field in the region outside
the coaxial cable and why?
r
I
Answer: H = 0 outside the coaxial cable because the net
current enclosed by an Ampèrian contour enclosing the
cable is zero.
a
b
C
I
Exercise 5-11: The metal niobium becomes a supercon-
ductor with zero electrical resistance when it is cooled to
below 9 K, but its superconductive behavior ceases when
the magnetic flux density at its surface exceeds 0.12 T.
Determine the maximum current that a 0.1-mm diameter
niobium wire can carry and remain superconductive.
Answer: I = 30 A. (See
Example 5-5:
EM
Figure 5-18 Toroidal coil with inner radius a and outer
radius b. The wire loops usually are much more closely spaced
than shown in the figure (Example 5-5).
.)
Magnetic Field inside a
Toroidal Coil
A toroidal coil (also called a torus or toroid) is a doughnutshaped structure (called the core) wrapped in closely spaced
turns of wire (Fig. 5-18). For clarity, we show the turns in the
figure as spaced far apart, but in practice, they are wound in
a closely spaced arrangement to form approximately circular
loops. The toroid is used to magnetically couple multiple
circuits and to measure the magnetic properties of materials, as
illustrated later in Fig. 5-31. For a toroid with N turns carrying
a current I, determine the magnetic field H in each of the
following three regions: r < a, a < r < b, and r > b with all in
the azimuthal symmetry plane of the toroid.
Solution: From symmetry, it is clear that H is uniform in
the azimuthal direction. If we construct a circular Ampèrian
contour with center at the origin and radius r < a, there will
be no current flowing through the surface of the contour.
Therefore,
H=0
for r < a.
Similarly, for an Ampèrian contour with radius r > b, the net
current flowing through its surface is zero because an equal
number of current coils cross the surface in both directions;
hence,
H=0
Ampèrian contour
for r > b (region exterior to the toroidal coil).
For the region inside the core, we construct a path of radius r
(Fig. 5-18). For each loop, we know from Example 5-3 that
the field H at the center of the loop points along the axis
of the loop, which in this case is the φ direction, and in
view of the direction of the current I shown in Fig. 5-18, the
right-hand rule tells us that H must be in the −φ direction.
Hence, H = −φ̂φH. The total current crossing the surface
of the contour with radius r is NI and its direction is into
the page. According to the right-hand rule associated with
Ampère’s law, the current is positive if it crosses the surface
of the contour in the direction of the four fingers of the right
hand when the thumb is pointing along the direction of the
contour C. Hence, the current through the surface spanned by
the contour is −NI. Application of Ampère’s law then gives
Z
C
H · dl =
Z 2π
0
(−φ̂φH) · φ̂φr d φ
= −2π rH
= −NI.
Hence, H = NI/(2π r) and
H = −φ̂φH = −φ̂φ
NI
2π r
(for a < r < b).
(5.50)
246
CHAPTER 5
◮ The magnetic field induced by a current-carrying coil
wound around a toroid is confined entirely to the toroid
volume:


for r < a,
0
H = −φ̂φ 2NI
for
a < r < b,
πr

0
for r > b.
Example 5-6:
Magnetic Field of an Infinite
Current Sheet
The x–y plane contains an infinite current sheet with surface
current density Js = x̂Js (Fig. 5-19). Find the magnetic field H
everywhere in space.
z
Ampèrian
contour
H
for z > 0,
(5.51)
for z < 0.
What are the fundamental differences between electric and magnetic fields?
Concept Question 5-8:
If the line integral of H over a
closed contour is zero, does it follow that H = 0 at every
point on the contour? If not, what then does it imply?
Concept Question 5-9:
Compare the utility of applying the Biot–Savart law versus applying Ampère’s law
for computing the magnetic field due to current-carrying
conductors.
Concept Question 5-10:
What is a toroid? What is the
magnetic field outside the toroid?
Concept Question 5-11:
l
y
w
H
Js (out of the page)
Figure 5-19 A thin current sheet in the x–y plane carrying a
surface current density Js = x̂Js (Example 5-6).
Solution: From symmetry considerations and the right-hand
rule, for z > 0 and z < 0 H must be in the directions shown in
the figure. That is,
−ŷH
for z > 0,
H=
ŷH
for z < 0.
To evaluate the line integral in Ampère’s law, we choose a
rectangular Ampèrian path around the sheet with dimensions
l and w (Fig. 5-19). Recalling that Js represents current per
unit length along the y direction, the total current crossing
the surface of the rectangular loop is I = Js l. Hence, applying
Ampère’s law over the loop in a counterclockwise direction,
while noting that H is perpendicular to the paths of length w,
we have
Z
H · dl = 2Hl = Js l,
C
from which we obtain the result

Js

 −ŷ
2
H=

 ŷ Js
2
MAGNETOSTATICS
5-4 Vector Magnetic Potential
In our treatment of electrostatic fields in Chapter 4, we defined
the electrostatic potential V as the line integral of the electric
field E and found that V and E are related by E = −∇V .
This relationship proved useful not only in relating electric
field distributions in circuit elements (such as resistors and
capacitors) to the voltages across them but also to determine
E for a given charge distribution by first computing V using
Eq. (4.48). We now explore a similar approach in connection
with the magnetic flux density B.
According to Eq. (5.44), ∇ · B = 0. We wish to define B in
terms of a magnetic potential with the constraint that such a
definition guarantees that the divergence of B is always zero.
This can be realized by taking advantage of the vector identity
given by Eq. (3.106b), which states that, for any vector A,
× A) = 0.
∇ ·(∇×
(5.52)
Hence, by introducing the vector magnetic potential A such
that
×A
B = ∇×
(Wb/m2 ),
(5.53)
we are guaranteed that ∇ · B = 0. The SI unit for B is the
tesla (T). An equivalent unit is webers per square meter
(Wb/m2 ). Consequently, the SI unit for A is (Wb/m).
5-4 VECTOR MAGNETIC POTENTIAL
247
Technology Brief 10: Electromagnets
Magnetic Relays
William Sturgeon developed the first practical electromagnet in the 1820s. Today, the principle of the electromagnet is used in motors, relay switches in read/write
heads for hard disks and tape drives, loud speakers,
magnetic levitation, and many other applications.
Basic Principle
A magnetic relay is a switch or circuit breaker that can
be activated into the “ON” and “OFF” positions magnetically. One example is the low-power reed relay used in
telephone equipment, which consists of two flat nickel–
iron blades separated by a small gap (Fig. TF10-2). The
blades are shaped in such a way that, in the absence of
an external force, they remain apart and unconnected
(OFF position). Electrical contact between the blades
Electromagnets can be constructed in various shapes,
including the linear solenoid and horseshoe geometries depicted in Fig. TF10-1. In both cases, when an
electric current flows through the insulated wire coiled
around the central core, it induces a magnetic field with
lines resembling those generated by a bar magnet. The
strength of the magnetic field is proportional to the current, the number of turns, and the magnetic permeability
of the core material. By using a ferromagnetic core,
the field strength can be increased by several orders of
magnitude, depending on the purity of the iron material.
When subjected to a magnetic field, ferromagnetic materials, such as iron or nickel, get magnetized and act like
magnets themselves.
Glass envelope
S
N
Electronic circuit
Figure TF10-2 Micro-reed relay (size exaggerated for illustration purposes).
N
Switch
Iron core
Iron core
B
B
N
S
B
Insulated wire
S
(a) Solenoid
Magnetic field
(b) Horseshoe electromagnet
Figure TF10-1 Solenoid and horseshoe magnets.
248
CHAPTER 5
(ON position) is realized by applying a magnetic field
along their length. The field, induced by a current flowing
in the wire coiled around the glass envelope, causes
the two blades to assume opposite magnetic polarities,
thereby forcing them to attract together and close out the
gap.
MAGNETOSTATICS
The Loudspeaker
By using a combination of a stationary, permanent magnet and a moveable electromagnet, the electromagnet/
speaker-cone of the loudspeaker (Fig. TF10-4) can be
made to move back and forth in response to the electrical
The Doorbell
In a doorbell circuit (Fig. TF10-3), the doorbell button
is a switch; pushing and holding it down serves to
connect the circuit to the household ac source through
an appropriate step-down transformer. The current
from the source flows through the electromagnet, via a
contact arm with only one end anchored in place (and
the other moveable), and onward to the switch. The
magnetic field generated by the current flowing in the
windings of the electromagnet pulls the unanchored end
of the contact arm (which has an iron bar on it) closer
in—in the direction of the electromagnet—thereby losing
connection with the metal contact and severing current
flow in the circuit. With no magnetic field to pull on the
contact arm, it snaps back into its earlier position, reestablishing the current in the circuit. This back-and-forth
cycle is repeated many times per second, so long as the
doorbell button continues to be pushed down, and with
every cycle, the clapper arm attached to the contact arm
hits the metal bell and generates a ringing sound.
Permanent
magnet
Cone
Electrical
signal
Figure TF10-4 The basic structure of a speaker.
Electromagnet
Magnetic field
Clapper
Bell
Audio
signal
Contact arm
ac source
Metal contact
Button
Transformer
Figure TF10-3 Basic elements of a doorbell.
5-4 VECTOR MAGNETIC POTENTIAL
249
S
N
(a)
(b)
N
S
S
N
N
S
S
N
N
S
S
N
N
S
(c)
Figure TF10-5 (a) A maglev train, (b) electrodynamic suspension of an SCMaglev train, and (c) electrodynamic maglev propulsion via
propulsion coils.
signal exciting the electromagnet. The vibrating movement of the cone generates sound waves with the same
distribution of frequencies as contained in the spectrum
of the electrical signal.
Magnetic Levitation
◮ Magnetically levitated trains (Fig. TF10-5(a)),
called maglevs for short, can achieve speeds as
high as 500 km/hr, primarily because there is no
friction between the train and the track. ◭
The train basically floats at a height of 1 or more
centimeters above the track, which is made possible by
magnetic levitation (Fig. TF10-5(b)). The train carries
superconducting electromagnets that induce currents in
coils built into the guide rails alongside the train. The
magnetic interaction between the train’s superconducting
electromagnets and the guide-rail coils serves not only to
levitate the train but also to propel it along the track.
250
CHAPTER 5
5-4.1 Vector Poisson’s Equation
With B = µ H, the differential form of Ampère’s law given by
Eq. (5.46) can be written as
× B = µ J.
∇×
(5.54)
If we substitute Eq. (5.53) into Eq. (5.54), we obtain
× (∇×
× A) = µ J.
∇×
(5.55)
and its solution for a volume charge distribution ρv occupying
a volume υ ′ was given by Eq. (4.71) as
1
V=
4πε
(5.56)
where, by definition, ∇2 A in Cartesian coordinates is
2
∂
∂2
∂2
2
A
∇ A=
+
+
∂ x2 ∂ y2 ∂ z2
= x̂∇2 Ax + ŷ∇2 Ay + ẑ∇2 Az .
µ
Ax =
4π
Combining Eq. (5.55) with Eq. (5.56) gives
∇(∇ · A) − ∇2 A = µ J.
(5.58)
This equation contains a term involving ∇ · A. It turns out
that we have a fair amount of latitude in specifying a value
or a mathematical form for ∇ · A without conflicting with the
requirement represented by Eq. (5.53). The simplest among
these allowed restrictions on A is
∇ · A = 0.
(5.60)
Using the definition for ∇2 A given by Eq. (5.57), the vector
Poisson’s equation can be decomposed into three scalar Poisson’s equations:
∇2 Ax = −µ Jx ,
∇2 Ay = −µ Jy ,
∇2 Az = −µ Jz .
(5.61a)
(5.61b)
ρv
,
ε
(5.63)
Z
Jx
dυ ′
R′
υ′
(Wb/m).
(5.64)
µ
4π
Z
υ′
J
dυ ′
R′
(Wb/m).
(5.65)
In view of Eq. (5.23), if the current distribution is specified
over a surface S′ , then J d υ ′ should be replaced with Js ds′ and
υ ′ should be replaced with S′ ; similarly, for a line distribution,
J d υ ′ should be replaced with I dl′ , and the integration should
be performed over the associated path l ′ .
◮ The vector magnetic potential provides a third approach
for computing the magnetic field due to current-carrying
conductors in addition to the methods suggested by the
Biot–Savart and Ampère laws. ◭
For a specified current distribution, Eq. (5.65) can be used to
find A, and then Eq. (5.53) can be used to find B. Except for
simple current distributions with symmetrical geometries that
lend themselves to the application of Ampère’s law, in practice
we often use the approaches provided by the Biot–Savart law
and the vector magnetic potential, and among these two, the
latter often is more convenient to apply because it is easier to
perform the integration in Eq. (5.65) than that in Eq. (5.22).
5-4.2
Magnetic Flux
The magnetic flux Φ linking a surface S is defined as the total
magnetic flux density passing through it, or
(5.61c)
In electrostatics, Poisson’s equation for the scalar potential V
is given by Eq. (4.70) as
∇2V = −
ρv
dυ ′.
R′
(5.59)
Using this choice in Eq. (5.58) leads to the vector Poisson’s
equation
∇2 A = −µ J.
υ′
Similar solutions can be written for Ay in terms of Jy and for
Az in terms of Jz . The three solutions can be combined into a
vector equation:
A=
(5.57)
Z
Poisson’s equations for Ax , Ay , and Az are mathematically
identical in form to Eq. (5.62). Hence, for a current density J
with x component Jx distributed over a volume υ ′ , the solution
for Eq. (5.61a) is
For any vector A, the Laplacian of A obeys the vector identity
given by Eq. (3.113), that is,
× (∇×
× A),
∇2 A = ∇(∇ · A) − ∇×
MAGNETOSTATICS
(5.62)
Φ=
Z
S
B · ds
(Wb).
(5.66)
If we insert Eq. (5.53) into Eq. (5.66) and then invoke Stokes’s
theorem, we obtain
Φ=
Z
S
(∇ × A) · ds =
Z
C
A · dl
(Wb),
(5.67)
5-5 MAGNETIC PROPERTIES OF MATERIALS
251
where C is the contour bounding the surface S. Thus, Φ can
be determined by either Eq. (5.66) or Eq. (5.67), whichever is
easier to integrate for the specific problem under consideration.
r
e
e
5-5 Magnetic Properties of Materials
Because of the similarity between the pattern of the magnetic
field lines generated by a current loop and those exhibited
by a permanent magnet, the loop can be regarded as a magnetic dipole with north and south poles (Section 5-2.2 and
Fig. 5-13). The magnetic moment m of a loop of area A has
magnitude m = IA and a direction normal to the plane of the
loop (in accordance with the right-hand rule). Magnetization in
a material is due to atomic scale current loops associated with
(1) orbital motions of the electrons and protons around and
inside the nucleus and (2) electron spin. The magnetic moment
due to proton motion typically is three orders of magnitude
smaller than that of the electrons; therefore, the total orbital
and spin magnetic moment of an atom is dominated by the
sum of the magnetic moments of its electrons.
◮ The magnetic behavior of a material is governed by the
interaction of the magnetic dipole moments of its atoms
with an external magnetic field. The nature of the behavior
depends on the crystalline structure of the material and is
used as a basis for classifying materials as diamagnetic,
paramagnetic, or ferromagnetic. ◭
The atoms of a diamagnetic material have no permanent magnetic moments. In contrast, both paramagnetic and ferromagnetic materials have atoms with permanent magnetic dipole
moments, albeit with very different organizational structures.
5-5.1 Electron Orbital and Spin Magnetic
Moments
This section presents a semiclassical, intuitive model of the
atom, which provides quantitative insight into the origin of
electron magnetic moments. An electron with charge of −e
moving at constant speed u in a circular orbit of radius r
(Fig. 5-20(a)) completes one revolution in time T = 2π r/u.
This circular motion of the electron constitutes a tiny loop with
current I given by
e
eu
I=− =−
.
(5.68)
T
2π r
The magnitude of the associated orbital magnetic moment mo
is
eu eur
e
mo = IA = −
Le , (5.69)
=−
(π r2 ) = −
2π r
2
2me
ms
mo
(a) Orbiting electron
(b) Spinning electron
Figure 5-20 An electron generates (a) an orbital magnetic
moment mo as it rotates around the nucleus and (b) a spin
magnetic moment ms as it spins about its own axis.
where Le = me ur is the angular momentum of the electron and
me is its mass. According to quantum physics, the orbital angular momentum is quantized; specifically, Le is always some
integer multiple of h̄ = h/2π , where h is Planck’s constant.
That is, Le = 0, h̄, 2h̄, . . . . Consequently, the smallest nonzero
magnitude of the orbital magnetic moment of an electron is
mo = −
eh̄
.
2me
(5.70)
Despite the fact that all materials contain electrons that exhibit
magnetic dipole moments, most are effectively nonmagnetic.
This is because, in the absence of an external magnetic field,
the atoms of most materials are oriented randomly, as a result
of which they exhibit a zero or very small net magnetic
moment.
In addition to the magnetic moment due to its orbital motion,
an electron has an intrinsic spin magnetic moment ms due
to its spinning motion about its own axis (Fig. 5-20(b)). The
magnitude of ms predicted by quantum theory is
ms = −
eℏ
,
2me
(5.71)
which is equal to the minimum orbital magnetic moment mo .
The electrons of an atom with an even number of electrons
usually exist in pairs with the members of a pair having
opposite spin directions, thereby canceling each others’ spin
magnetic moments. If the number of electrons is odd, the atom
has a net nonzero spin magnetic moment due to its unpaired
electron.
5-5.2 Magnetic Permeability
In Chapter 4, we learned that the relationship D = ε0 E,
between the electric flux and field in free space, is modified to
D = ε0 E + P in a dielectric material. Likewise, the relationship
252
CHAPTER 5
B = µ0 H in free space is modified to
Ferromagnetic materials are discussed next.
B = µ0 H + µ0M = µ0 (H + M),
(5.72)
where the magnetization vector M is defined as the vector sum
of the magnetic dipole moments of the atoms contained in a
unit volume of the material. Scale factors aside, the roles and
interpretations of B, H, and M in Eq. (5.72) mirror those of D,
E, and P in Eq. (4.93). Moreover, just as in most dielectrics P
and E are linearly related, in most magnetic materials
M = χm H,
(5.73)
where χm is a dimensionless quantity called the magnetic
susceptibility of the material. For diamagnetic and paramagnetic materials, χm is a (temperature-dependent) constant,
resulting in a linear relationship between M and H at a given
temperature. This is not the case for ferromagnetic substances;
the relationship between M and H not only is nonlinear, but
also depends on the “history” of the material, as explained in
the next section.
Keeping this fact in mind, we can combine Eqs. (5.72) and
(5.73) to get
B = µ0 (H + χm H) = µ0 (1 + χm )H,
(5.74)
B = µ H,
(5.75)
or
where µ , the magnetic permeability of the material, relates
to χm as
µ = µ0 (1 + χm)
(H/m).
(5.76)
Often it is convenient to define the magnetic properties of a
material in terms of the relative permeability µr :
µr =
MAGNETOSTATICS
µ
= 1 + χm .
µ0
(5.77)
A material usually is classified as diamagnetic, paramagnetic,
or ferromagnetic on the basis of the value of its χm (Table 5-2).
Diamagnetic materials have negative susceptibilities, whereas
paramagnetic materials have positive ones. However, the absolute magnitude of χm is on the order of 10−5 for both classes
of materials, which for most applications allows us to ignore
χm relative to 1 in Eq. (5.77).
◮ Thus, µr ≈ 1 or µ ≈ µ0 for diamagnetic and paramagnetic substances, which include dielectric materials
and most metals. In contrast, |µr | ≫ 1 for ferromagnetic
materials; |µr | of purified iron, for example, is on the order
of 2 × 105. ◭
Exercise 5-12: The magnetic vector M is the vector
sum of the magnetic moments of all the atoms contained
in a unit volume (1m3 ). If a certain type of iron with
8.5 × 1028 atoms/m3 contributes one electron per atom
to align its spin magnetic moment along the direction of
the applied field, find (a) the spin magnetic moment of
a single electron given that me = 9.1 × 10−31 (kg) and
h̄ = 1.06 × 10−34 (J·s) and (b) the magnitude of M.
Answer: (a) ms = 9.3 × 10−24 (A·m2 ), (b) M = 7.9 × 105
(A/m). (See
5-5.3
EM
.)
Magnetic Hysteresis of Ferromagnetic
Materials
Ferromagnetic materials, which include iron, nickel, and
cobalt, exhibit unique magnetic properties due to the fact that
their magnetic moments tend to readily align along the direction of an external magnetic field. Moreover, such materials
remain partially magnetized even after the external field is
removed. Because of these peculiar properties, ferromagnetic
materials are used in the fabrication of permanent magnets.
A key to understanding the properties of ferromagnetic
materials is the notion of magnetized domains, that are microscopic regions (on the order of 10−10 m3 ) within which
the magnetic moments of all atoms (typically on the order
of 1019 atoms) are permanently aligned with each other. This
alignment, which occurs in all ferromagnetic materials, is
due to strong coupling forces between the magnetic dipole
moments constituting an individual domain. In the absence of
an external magnetic field, the domains take on random orientations relative to each other (Fig. 5-21(a)), resulting in zero
net magnetization. The domain walls forming the boundaries
between adjacent domains consist of thin transition regions.
When an unmagnetized sample of a ferromagnetic material
is placed in an external magnetic field, the domains partially
align with the external field, as illustrated in Fig. 5-21(b). A
quantitative understanding of how the domains form and how
they behave under the influence of an external magnetic field
requires a heavy dose of quantum mechanics and is outside
the scope of the present treatment. Hence, we confine our
discussion to a qualitative description of the magnetization
process and its implications.
The magnetization behavior of a ferromagnetic material is
described in terms of its B–H magnetization curve, where B
and H refer to the amplitudes of the B flux and H field in the
material. Suppose that we start with an unmagnetized sample
of iron, denoted by point O in Fig. 5-22. When we increase H
continuously, for example, by increasing the current passing
5-5 MAGNETIC PROPERTIES OF MATERIALS
253
Table 5-2 Properties of magnetic materials.
Diamagnetism
Paramagnetism
Ferromagnetism
No
Yes, but weak
Yes, and strong
Electron orbital
magnetic moment
Electron spin
magnetic moment
Magnetized
domains
Opposite
Same
Hysteresis
[see Fig. 5-22]
Common substances
Bismuth, copper, diamond,
gold, lead, mercury, silver,
silicon
Aluminum, calcium,
chromium, magnesium,
niobium, platinum,
tungsten
Iron,
nickel,
cobalt
Typical value of χm
Typical value of µr
≈ −10−5
≈1
≈ 10−5
≈1
|χm | ≫ 1 and hysteretic
|µr | ≫ 1 and hysteretic
Permanent magnetic
dipole moment
Primary magnetization
mechanism
Direction of induced
magnetic field
(relative to external field)
B
A1
A2
A3
O
Br
H
(a) Unmagnetized domains
A4
Figure 5-22 Typical hysteresis curve for a ferromagnetic
material.
(b) Magnetized domains
Figure 5-21 Comparison of (a) unmagnetized and (b) magnetized domains in a ferromagnetic material.
through a wire wound around the sample, B increases also
along the B–H curve from point O to point A1 , at which
nearly all the domains have become aligned with H. Point A1
represents a saturation condition. If we then decrease H from
its value at point A1 back to zero (by reducing the current
through the wire), the magnetization curve follows the path
from A1 to A2 . At point A2 , the external field H is zero (owing
to the fact that the current through the wire is zero), but the
flux density B in the material is not. The magnitude of B at A2
is called the residual flux density Br . The iron material is now
magnetized and ready to be used as a permanent magnet owing
to the fact that a large fraction of its magnetized domains have
remained aligned. Reversing the direction of H and increasing
its intensity causes B to decrease from Br at point A2 to zero
at point A3 , and if the intensity of H is increased further
254
CHAPTER 5
B
(a) Hard material
What does a magnetization
curve describe? What is the difference between the magnetization curves of hard and soft ferromagnetic materials?
Concept Question 5-14:
B
H
H
(b) Soft material
Figure 5-23 Comparison of hysteresis curves for (a) a hard
ferromagnetic material and (b) a soft ferromagnetic material.
while maintaining its direction, the magnetization moves to
the saturation condition at point A4 . Finally, as H is made
to return to zero and is then increased again in the positive
direction, the curve follows the path from A4 to A1 . This
process is called magnetic hysteresis. Hysteresis means “lag
behind.” The existence of a hysteresis loop implies that the
magnetization process in ferromagnetic materials depends not
only on the magnetic field H but also on the magnetic history
of the material. The shape and extent of the hysteresis loop
depend on the properties of the ferromagnetic material and
the peak-to-peak range over which H is made to vary. Hard
ferromagnetic materials are characterized by wide hysteresis
loops (Fig. 5-23(a)). They cannot be easily demagnetized by
an external magnetic field because they have a large residual
magnetization Br . Hard ferromagnetic materials are used in the
fabrication of permanent magnets for motors and generators.
Soft ferromagnetic materials have narrow hysteresis loops
(Fig. 5-23(b)); hence, they can be more easily magnetized and
demagnetized. To demagnetize any ferromagnetic material,
the material is subjected to several hysteresis cycles while
gradually decreasing the peak-to-peak range of the applied
field.
What are the three types of
magnetic materials, and what are typical values of their
relative permeabilities?
Concept Question 5-12:
What causes magnetic hysteresis in ferromagnetic materials?
Concept Question 5-13:
MAGNETOSTATICS
5-6 Magnetic Boundary Conditions
In Chapter 4, we derived a set of boundary conditions that
describes how, at the boundary between two dissimilar contiguous media, the electric flux and field D and E in the first
medium relate to those in the second medium. We now derive
a similar set of boundary conditions for the magnetic flux
and field B and H. By applying Gauss’s law to a pill box
that straddles the boundary, we determined that the difference
between the normal components of the electric flux densities
in two media equals the surface charge density ρs . That is,
Z
S
D · ds = Q
D1 n − D2 n = ρ s .
(5.78)
By analogy, application of Gauss’s law for magnetism, as
expressed by Eq. (5.44), leads to the conclusion that
Z
S
B · ds = 0
B1n = B2n .
(5.79)
◮ Thus, the normal component of B is continuous across
the boundary between two adjacent media. ◭
Because B1 = µ1 H1 and B2 = µ2 H2 for linear, isotropic
media, the boundary condition for H corresponding to
Eq. (5.79) is
µ1 H1n = µ2 H2n .
(5.80)
Comparison of Eqs. (5.78) and (5.79) reveals a striking
difference between the behavior of the magnetic and electric
fluxes across a boundary: Whereas the normal component of B
is continuous across the boundary, the normal component of D
is not (unless ρs = 0). The reverse applies to the tangential
components of the electric and magnetic fields E and H:
Whereas the tangential component of E is continuous across
the boundary, the tangential component of H is not (unless
the surface current density Js = 0). To obtain the boundary
condition for the tangential component of H, we follow the
same basic procedure used in Chapter 4 to establish the
boundary condition for the tangential component of E. With
reference to Fig. 5-24, we apply Ampère’s law (Eq. (5.47)) to
5-6 MAGNETIC BOUNDARY CONDITIONS
H1
H1n
b
255
l̂l 1
n̂
H1t
H2n
l̂l 2
H2t
c
H2
∆l
a
}
}
d
Medium 1
μ1
nˆ 2
∆h
2
∆h
2
Js
Medium 2
μ2
Figure 5-24 Boundary between medium 1 with µ1 and medium 2 with µ2 .
a closed rectangular path with sides of lengths ∆l and ∆h, and
then let ∆h → 0 to obtain
Z
C
H · dl =
Z b
a
H1 · ℓˆ1 dℓ +
Z d
c
H2 · ℓˆ2 dℓ = I,
(5.81)
where I is the net current crossing the surface of the loop in the
direction specified by the right-hand rule (I is in the direction
of the thumb when the fingers of the right hand extend in the
direction of the loop C). As we let ∆h of the loop approach
zero, the surface of the loop approaches a thin line of length ∆l.
The total current flowing through this thin line is I = Js ∆l,
where Js is the magnitude of the component of the surface
current density Js normal to the loop. That is, Js = Js · n̂, where
n̂ is the normal to the loop. In view of these considerations,
Eq. (5.81) becomes
(H1 − H2) · ℓˆ1 ∆l = Js · n̂ ∆l.
Example 5-7: Tilted-Plane Boundary
In Fig. 5-25 the plane defined by y + 2x = 2 separates
medium 1 of permeability µ1 from medium 2 of permeability
µ2 . If no surface current exists on the boundary and
B1 = x̂2 + ŷ3
(T),
find B2 and then evaluate the result for µ2 = 2µ1 .
y
(5.82)
The vector ℓˆ1 can be expressed as ℓˆ1 = n̂ × n̂2 , where n̂ and
n̂2 are the normals to the loop and to the surface of medium 2
(Fig. 5-24), respectively. Using this relation in Eq. (5.82) and
× C) = B ·(C×
× A) leads
then applying the vector identity A ·(B×
to
n̂ ·[n̂2 × (H1 − H2 )] = Js · n̂.
(5.83)
(0, 2)
Medium 1
μ1
θ
(1, 0)
x
Since Eq. (5.83) is valid for any n̂, it follows that
n̂2 × (H1 − H2 ) = Js .
This equation implies that the tangential components of H
parallel to Js are continuous across the interface, whereas those
orthogonal to Js are discontinuous in the amount of Js .
Surface currents can exist only on the surfaces of perfect conductors and superconductors. Hence, at the interface
between media with finite conductivities, Js = 0 and
H1t = H2t .
n2y
(5.84)
(5.85)
Medium 2
μ2
nˆ 2
θ
n2x
Plane y + 2x = 2
Figure 5-25 Magnetic media separated by the plane y + 2x = 2
(Example 5-7).
256
CHAPTER 5
Solution: To apply the boundary conditions at the plane
between the two media, we first need to obtain an expression
for the unit vector normal to the boundary. To apply the
boundary condition given by Eq. (5.84), we need an expression
for n̂2 , which isthe unit vector pointing away from medium 2
(Fig. 5-25).
From the geometry of the small triangle in Fig. 5-25,
θ = tan−1
which leads to
µ2
B1t
µ1
µ2
B1t
= B1n n̂2 +
µ1
B2 = B2n + B2t = B1n +
= (2 cos θ + 3 sin θ )(x̂ cos θ + ŷsin θ )
µ2
+
[x̂(2 − 2 cos2 θ − 3 sin θ cos θ )
µ1
1
2
+ ŷ(3 − 2 cos θ sin θ − 3 sin2 θ )]
= x̂ 2 cos2 θ + 3 sin θ cos θ
= 26.57◦.
Hence,
µ2
2
(2 − 2 cos θ − 3 sin θ cos θ )
+
µ1
+ ŷ 2 cos θ sin θ + 3 sin2 θ
n̂2 = x̂ cos θ + ŷ sin θ
= x̂0.89 + ŷ0.45.
Magnetic field B1 has a normal component B1n along n̂2 and a
tangential component B1t :
B1 = B1n + B1t
MAGNETOSTATICS
µ2
2
+
(3 − 2 cos θ sin θ − 3 sin θ ) .
µ1
For θ = 26.57◦ and µ2 = 2µ1 ,
with
B2 = x̂1.2 + ŷ4.6
(T).
B1n = B1n n̂2
Exercise 5-13: With reference to Fig. 5-24, determine the
and
B1n = n̂2 · B1
= (x̂ cos θ + ŷ sin θ ) ·(x̂2 + ŷ3)
= 2 cos θ + 3 sin θ .
The tangential component of B1 is
B1t = B1 − B1n
= B1 − B1n n̂2
= (x̂2 + ŷ3) − (2 cos θ + 3 sin θ )(x̂ cos θ + ŷ sin θ )
= x̂(2 − 2 cos2 θ − 3 sin θ cos θ )
+ ŷ(3 − 2 cos θ sin θ − 3 sin2 θ ).
Boundary conditions require that
angle between H1 and n̂2 = ẑ if H2 = (x̂3 + ẑ2) (A/m),
µr1 = 2, µr2 = 8, and Js = 0.
Answer: θ = 20.6◦ . (See
EM
.)
5-7 Inductance
An inductor is the magnetic analogue of an electric capacitor.
Just as a capacitor can store energy in the electric field in
the medium between its conducting surfaces, an inductor can
store energy in the magnetic field near its current-carrying conductors. A typical inductor consists of multiple turns of wire
helically coiled around a cylindrical core (Fig. 5-26(a)). Such
a structure is called a solenoid. Its core may be air filled or may
contain a magnetic material with magnetic permeability µ . If
the wire carries a current I and the turns are closely spaced,
the solenoid will produce a relatively uniform magnetic field
within its interior with magnetic field lines resembling those
of the permanent magnet (Fig. 5-26(b)).
B1n = B2n
5-7.1
and
B1t
B2t
=
,
µ1
µ2
Magnetic Field in a Solenoid
As a prelude to our discussion of inductance, we derive an
expression for the magnetic flux density B in the interior
5-7 INDUCTANCE
257
z
B
N
a
N
dz
B
B
dθ
θ
P
l
z
θ2
x
θ1
S
S
I (out)
(a) Loosely wound
solenoid
(b) Tightly wound
solenoid
I (in)
(a) Cross section
Figure 5-26 Magnetic field lines of (a) a loosely wound
solenoid and (b) a tightly wound solenoid.
region of a tightly wound solenoid. The solenoid is of length l
and radius a and comprises N turns carrying current I. The
number of turns per unit length is n = N/l, and the fact that
the turns are tightly wound implies that the pitch of a single
turn is small compared with the solenoid’s radius. Even though
the turns are slightly helical in shape, we can treat them as
circular loops (Fig. 5-27(a)). Let us start by considering the
magnetic flux density B at point P on the axis of the solenoid.
In Example 5-3, we derived the following expression for the
magnetic field H along the axis of a circular loop of radius a,
which is a distance z away from its center:
H = ẑ
I ′ a2
2(a2 + z2 )3/2
L=μ
N2
S
l
S
Cross section
(b) Solenoid inductance
,
(5.86)
I′
where
is the current carried by the loop. If we treat an
incremental length dz of the solenoid as an equivalent loop
composed of n dz turns carrying a current I ′ = In dz, then the
induced field at point P is
µ nIa2
dz.
2(a2 + z2 )3/2
Figure 5-27 (a) Solenoid cross section showing geometry for
calculating H at a point P on the solenoid axis and (b) solenoid
inductance.
by expressing the variable z in terms of the angle θ , as seen
from P to a point on the solenoid rim. That is,
dB = µ dH
= ẑ
l
N
(5.87)
The total field B at P is obtained by integrating the contributions from the entire length of the solenoid. This is facilitated
z = a tan θ ,
a2 + z2 = a2 + a2 tan2 θ = a2 sec2 θ ,
2
dz = a sec θ d θ .
(5.88a)
(5.88b)
(5.88c)
258
CHAPTER 5
Upon substituting the last two expressions in Eq. (5.87) and
integrating from θ1 to θ2 , we have
µ nIa2
2
Z θ2
a sec2 θ d θ
µ nI
(sin θ2 − sin θ1 ).
2
(5.89)
If the solenoid length l is much larger than its radius a, then
for points P away from the solenoid’s ends θ1 ≈ −90◦ and
θ2 ≈ 90◦, in which case Eq. (5.89) reduces to
B = ẑ
θ1
a3 sec3 θ
ẑµ NI
l
(long solenoid with l/a ≫ 1).
Exercise 5-14: Use Eq. (5.89) to obtain an expression for
B at a point on the axis of a very long solenoid but situated
at its end points. How does B at the end points compare to
B at the midpoint of the solenoid?
Answer: B = ẑ(µ NI/2l) at the end points, which is half
as large as B at the midpoint. (See EM .)
5-7.2 Self-Inductance of a Solenoid
From Eq. (5.66), the magnetic flux Φ linking a surface S is
Φ=
S
B · ds
(Wb).
Λ = NΦ = µ
(5.91)
In a solenoid characterized by an approximately uniform magnetic field throughout its cross-section and given by Eq. (5.90),
the flux linking a single loop is
Z N
N
Φ = ẑ µ I · ẑ ds = µ IS,
(5.92)
l
l
S
where S is the cross-sectional area of the loop (Fig. 5-27(b)).
Magnetic flux linkage Λ is defined as the total magnetic flux
linking a given circuit or conducting structure. If the structure
N2
IS
l
(Wb).
(5.93)
The self-inductance of any conducting structure is defined
as the ratio of the magnetic flux linkage Λ to the current I
flowing through the structure:
(5.90)
Even though Eq. (5.90) was derived for the field B at the
midpoint of the solenoid, it is approximately valid everywhere
in the solenoid’s interior, except near the ends.
We now return to a discussion of inductance, which includes
the notion of self-inductance, representing the magnetic flux
linkage of a coil or circuit with itself, and mutual inductance,
which involves the magnetic flux linkage in a circuit due to the
magnetic field generated by a current in another one. Usually,
when the term inductance is used, the intended reference is to
self-inductance.
Z
consists of a single conductor with multiple loops, as in the
case of the solenoid, Λ equals the flux linking all loops of the
structure. For a solenoid with N turns,
= ẑ
B ≈ ẑµ nI =
MAGNETOSTATICS
L=
Λ
I
(H).
(5.94)
The SI unit for inductance is the henry (H), which is equivalent
to webers per ampere (Wb/A).
For a solenoid, use of Eq. (5.93) gives
L=µ
5-7.3
N2
S
l
(solenoid).
(5.95)
Self-Inductance of Other Conductors
The solenoid consists of a single inductor of cross section S.
If, on the other hand, the structure consists of two separate
conductors, as in the case of the parallel-wire and coaxial
transmission lines shown in Fig. 5-28, the flux linkage Λ
associated with a length l of either line refers to the flux Φ
through a closed surface between the two conductors, such as
the shaded areas in Fig. 5-28. In reality, there is also some
magnetic flux that passes through the conductors themselves,
but it may be ignored by assuming that currents flow only on
the surfaces of the conductors, in which case the magnetic field
inside the conductors vanishes. This assumption is justified
by the fact that our interest in calculating Λ is for the purpose of determining the inductance of a given structure, and
inductance is of interest primarily in the ac case (i.e., timevarying currents, voltages, and fields). As we will see later
in Section 7-5, the current flowing in a conductor under ac
conditions is concentrated within a very thin layer on the skin
of the conductor.
◮ For the parallel-wire transmission line, ac currents flow
on the outer surfaces of the wires, and for the coaxial
line, the current flows on the outer surface of the inner
conductor and on the inner surface of the outer one (the
current-carrying surfaces are those adjacent to the electric
and magnetic fields present in the region between the
conductors). ◭
5-7 INDUCTANCE
259
y
z
I
I
S
I
I
l
Radius a
μ
x
l
S
a
r
b
d
z
Outer
conductor
(a) Parallel-wire transmission line
Inner
conductor
I
S
a
l
I
b
Outer
conductor
Figure 5-29 Cross-sectional view of coaxial transmission line
and
denote H field out of and into the
(Example 5-8).
page, respectively.
c
conductors. It is given by Eq. (5.30) as
(b) Coaxial transmission line
Figure 5-28 To compute the inductance per unit length of
a two-conductor transmission line, we need to determine the
magnetic flux through the area S between the conductors.
For a two-conductor configuration similar to those of
Fig. 5-28, the inductance is given by
φ
B = φ̂
Example 5-8:
Z
S
Z b
B dr = l
(5.96)
Inductance of a Coaxial
Transmission Line
Z b
µI
a
a
B · ds.
µI
,
2π r
(5.97)
where r is the radial distance from the axis of the coaxial line.
Consider a transmission-line segment of length l as shown in
Fig. 5-29. Because B is perpendicular to the planar surface S
between the conductors, the flux through S is
Φ=l
Λ Φ 1
L= = =
I
I
I
I
b
µ Il
ln
dr =
.
2π r
2π
a
(5.98)
Using Eq. (5.96), the inductance per unit length of the coaxial
transmission line is
L′ =
µ
L Φ
b
.
= =
ln
l
lI
2π
a
(5.99)
5-7.4 Mutual Inductance
Develop an expression for the inductance per unit length of
a coaxial transmission line with inner and outer conductors
of radii a and b (Fig. 5-29) and an insulating material of
permeability µ .
Solution: The current I in the inner conductor generates
a magnetic field B throughout the region between the two
Magnetic coupling between two different conducting structures is described in terms of the mutual inductance between
them. For simplicity, consider the case of two multiturn closed
loops with surfaces S1 and S2 . Current I1 flows through the first
loop (Fig. 5-30), and no current flows through the second one.
The magnetic field B1 generated by I1 results in a flux Φ12
260
CHAPTER 5
MAGNETOSTATICS
I1
R
B1
V2
S2
I1
+
V1
−
N2 turns
S1
C2
C1
Figure 5-31
Toroidal coil with two windings used as a
transformer.
N1 turns
Figure 5-30 Magnetic field lines generated by current I1 in
z
loop 1 linking surface S2 of loop 2.
40 cm
through loop 2 given by
Φ12 =
Z
S2
B1 · ds,
20 A
(5.100)
and if loop 2 consists of N2 turns all coupled by B1 in exactly
the same way, then the total magnetic flux linkage through
loop 2 is
Z
Λ12 = N2 Φ12 = N2
S2
B1 · ds.
L12 =
Λ12
N2
=
I1
I1
Z
S2
B1 · ds
(H).
(5.102)
Mutual inductance is important in transformers (as discussed
in Chapter 6) wherein the windings of two or more circuits
share a common magnetic core, as illustrated by the toroidal
arrangement shown in Fig. 5-31.
Example 5-9:
5 cm
10 cm
20 cm
(5.101)
The mutual inductance associated with this magnetic coupling
is given by
Mutual Inductance
The rectangular conducting loop shown in Fig. 5-32 is coplanar with the long, straight wire carrying current I. Determine
the mutual inductance between the wire and the loop.
Solution: According to Eq. (5.102), the mutual inductance is
given by
Z
N2
L12 =
B1 · ds.
I1 S2
30 cm
y
x
Figure 5-32
Mutual inductance between two conductors
(Example 5-9).
In the present case, I1 = I is the current carried by the first
conductor (straight wire), S2 is the cross-sectional area of the
second conductor (rectangular loop), and B1 is the magnetic
field due to the first conductor (wire). Also, N2 = 1 because
the loop has a single turn. In the right-hand side of the
y–z plane, the magnetic field due to current I is into the page
(−x̂ direction) and given by
B = −x̂
µ0 I
,
2π y
where y is the distance away from the wire. The expression is
identical with that given by Eq. (5.97), after replacing φ̂φ with
−x̂ and r with y. To compute the magnetic flux passing through
the loop (into the page), we need to define the differential area
5-8 MAGNETIC ENERGY
261
as ds = −x̂ dy dz. Hence,
Z 0.4 Z
1 0.20
µ0 I
· (−x̂ dy dz)
2
πy
y=0.05 z=0.1
Z 0.20 Z 0.4 dy
µ0
dz
=
2π
0.1
0.05 y
0.3µ0
0.2
0.3 µ0
0.20
= 83 nH.
(ln y)|0.05 =
ln
=
2π
2π
0.05
L12 =
−x̂
I
What is the magnetic field
like in the interior of a long solenoid?
Concept Question 5-15:
What is the difference
between self-inductance and mutual inductance?
where υ = lS is the volume of the interior of the solenoid
and H = B/µ . The expression for Wm suggests that the energy
expended in building up the current in the inductor is stored in
the magnetic field with magnetic energy density wm , defined
as the magnetic energy Wm per unit volume,
wm =
Wm
υ
1
= µH2
2
(J/m3 ).
(5.105)
◮ Even though this expression was derived for a solenoid,
it remains valid for any medium with a magnetic
field H. ◭
Concept Question 5-16:
How is the inductance of a
solenoid related to its number of turns N?
Concept Question 5-17:
Furthermore, for any volume υ containing a material with
permeability µ (including free space with permeability µ0 ),
the total magnetic energy stored in a magnetic field H is
Wm =
1
2
Z
υ
µ H 2 dυ
(J).
(5.106)
5-8 Magnetic Energy
When we introduced electrostatic energy in Section 4-10, we
did so by examining what happens to the energy expended
in charging up a capacitor from zero voltage to some final
voltage V . We introduce the concept of magnetic energy by
considering an inductor with inductance L connected to a
current source. Suppose that we were to increase the current i
flowing through the inductor from zero to a final value I. From
circuit theory, we know that the instantaneous voltage υ across
the inductor is given by υ = L di/dt. We will derive this
relationship from Maxwell’s equations in Chapter 6, thereby
justifying the use of the i–υ relationship for the inductor.
Power p equals the product of υ and i, and the time integral of
power is work, or energy. Hence, the total energy in joules (J)
expended in building up a current I in the inductor is
Wm =
Z
p dt =
Z
iv dt = L
Z I
0
i di =
1 2
LI
2
(J).
(5.103)
We call this the magnetic energy stored in the inductor.
To justify this association, consider the solenoid inductor.
Its inductance is given by Eq. (5.95) as L = µ N 2 S/l, and the
magnitude of the magnetic flux density in its interior is given
by Eq. (5.90) as B = µ NI/l, implying that I = Bl/(µ N). Using
these expressions for L and I in Eq. (5.103), we obtain
1 2 1
Bl 2
N2
S
Wm = LI =
µ
2
2
l
µN
=
1 B2
1
(lS) = µ H 2 υ ,
2 µ
2
(5.104)
Example 5-10:
Magnetic Energy in a
Coaxial Cable
Derive an expression for the magnetic energy stored in a
coaxial cable of length l and inner and outer radii a and b.
The current flowing through the cable is I and its insulation
material has permeability µ .
Solution: From Eq. (5.97), the magnitude of the magnetic
field in the insulating material is
H=
B
I
=
,
µ
2π r
where r is the radial distance from the center of the inner
conductor (Fig. 5-29). The magnetic energy stored in the
coaxial cable therefore is
Wm =
1
2
Z
υ
µ H 2 dυ =
µ I2
8π 2
Z
υ
1
dυ .
r2
Since H is a function of r only, we choose d υ to be a
cylindrical shell of length l, radius r, and thickness dr along
the radial direction. Thus, d υ = 2π rl dr and
Z
1
b
µ I2 b 1
µ I 2l
= LI 2 (J)
Wm = 2
ln
· 2π rl dr =
2
8π a r
4π
a
2
with L given by Eq. (5.99).
262
CHAPTER 5
MAGNETOSTATICS
Technology Brief 11: Inductive Sensors
Linear Variable Differential Transformer (LVDT)
◮ An LVDT comprises a primary coil connected to
an ac source (typically a sine wave at a frequency
in the 1–10 kHz range) and a pair of secondary
coils, all sharing a common ferromagnetic core
(Fig. TF11-1). ◭
The magnetic core serves to couple the magnetic flux
generated by the primary coil into the two secondaries,
thereby inducing an output voltage across each of them.
The secondary coils are connected in opposition, so
when the core is positioned at the magnetic center of
the LVDT, the individual output signals of the secondaries
cancel each other out, producing a null output voltage.
The core is connected to the outside world via a nonmagnetic push rod. When the rod moves the core away
from the magnetic center, the magnetic fluxes induced
Vin
_
Primary coil
Phase
Amplitude and phase output
Magnetic coupling between different coils forms the
basis of several different types of inductive sensors.
Applications include the measurement of position and
displacement (with submillimeter resolution) in device
fabrication processes, proximity detection of conductive objects, and other related applications.
Amplitude
−10
−5
0
5
Distance traveled
10
Figure TF11-2 Amplitude and phase responses as functions
of the distance by which the magnetic core is moved away from
the center position.
in the secondary coils are no longer equal, resulting in
a nonzero output voltage. The LVDT is called a “linear”
transformer because the amplitude of the output voltage
is a linear function of displacement over a wide operating
range (Fig. TF11-2).
The cutaway view of the LVDT model in Fig. TF11-3
depicts a configuration in which all three coils—with
the primary straddled by the secondaries—are wound
around a glass tube that contains the magnetic core
and attached rod. Sample applications are illustrated in
Fig. TF11-4.
+
Stainless steel housing
Push
rod
Ferromagnetic core
Rod
Magnetic core
Secondary coils
_ V
out +
Figure TF11-1
(LVDT) circuit.
Primary coil
Secondary coils
Electronics
module
Linear variable differential transformer
Figure TF11-3 Cutaway view of LVDT.
5-8 MAGNETIC ENERGY
263
Sagging beam
LVDT
_
LVDT
Float
Primary
coil
Vin
+
Figure TF11-4 LVDT for measuring beam deflection and as
a fluid-level gauge.
_
Eddy-Current Proximity Sensor
Sensing
coil
Vout
+
The transformer principle can be applied to build a proximity sensor in which the output voltage of the secondary
coil becomes a sensitive indicator of the presence of a
conductive object in its immediate vicinity (Fig. TF11-5).
◮ When an object is placed in front of the secondary
coil, the magnetic field of the coil induces eddy (circular) currents in the object that generate magnetic
fields of their own having a direction that opposes the
magnetic field of the secondary coil. ◭
Eddy
currents
Conductive object
The reduction in magnetic flux causes a drop in output
voltage with the magnitude of the change being dependent on the conductive properties of the object and its
distance from the sensor.
Figure TF11-5 Eddy-current proximity sensor.
264
CHAPTER 5
MAGNETOSTATICS
Chapter 5 Summary
Concepts
• The magnetic force acting on a charged particle q
moving with a velocity u in a region containing a
× B.
magnetic flux density B is Fm = qu×
• The total electromagnetic force, known as the Lorentz
force, acting on a moving charge in the presence of both
× B).
electric and magnetic fields is F = q(E + u×
• Magnetic forces acting on current loops can generate
magnetic torques.
• The magnetic field intensity induced by a current element is defined by the Biot–Savart law.
• Gauss’s law for magnetism states that the net magnetic
flux flowing out of any closed surface is zero.
• Ampère’s law states that the line integral of H over a
closed contour is equal to the net current crossing the
surface bounded by the contour.
• The vector magnetic potential A is related to B by
× A.
B = ∇×
• Materials are classified as diamagnetic, paramagnetic,
•
•
•
•
•
or ferromagnetic, depending on their crystalline
structure and the behavior under the influence of an
external magnetic field.
Diamagnetic and paramagnetic materials exhibit a
linear behavior between B and H with µ ≈ µ0 for both.
Ferromagnetic materials exhibit a nonlinear hysteretic
behavior between B and H, and for some, µ may be as
large as 105 µ0 .
At the boundary between two different media, the normal component of B is continuous, and the tangential
components of H are related by H2t − H1t = Js , where
Js is the surface current density flowing in a direction
orthogonal to H1t and H2t .
The inductance of a circuit is defined as the ratio of
magnetic flux linking the circuit to the current flowing
through it.
Magnetic energy density is given by wm = 12 µ H 2 .
Mathematical and Physical Models
Maxwell’s Magnetostatics Equations
Gauss’s Law for Magnetism
Z
∇·B = 0
S
Ampère’s Law
B · ds = 0
Z
×H = J
∇×
C
H · dℓℓ = I
Magnetic Field
Infinitely Long Wire
Solenoid
Lorentz Force on Charge q
×A
B = ∇×
Magnetic Force on Wire
Z
C
×B
dl×
(N)
m = n̂ NIA
(N·m)
B ≈ ẑ µ nI =
ẑ µ NI
l
(Wb/m2 )
Vector Poisson’s Equation
∇2 A = −µ J
L=
(A·m2 )
Biot–Savart Law
Z
× R̂
I
dl×
H=
4 π l R2
Ia2
2(a2 + z2 )3/2
Inductance
Magnetic Torque on Loop
×B
T = m×
µ0 I
2π r
Vector Magnetic Potential
× B)
F = q(E + u×
Fm = I
H = ẑ
Circular Loop
B = φ̂φ
Λ Φ 1
= =
I
I
I
Z
S
B · ds
Magnetic Energy Density
(A/m)
wm =
1
µH2
2
(J/m3 )
(H)
(Wb/m2 )
(A/m)
(Wb/m2 )
PROBLEMS
Important Terms
265
Provide definitions or explain the meaning of the following terms:
Ampère’s law
Ampèrian contour
Biot–Savart law
current density (volume) J
diamagnetic
ferromagnetic
Gauss’s law for magnetism
hard and soft ferromagnetic
materials
inductance (self- and mutual)
Lorentz force F
magnetic dipole
magnetic energy Wm
magnetic energy
density wm
magnetic flux Φ
magnetic flux density B
magnetic flux linkage Λ
magnetic force Fm
magnetic hysteresis
magnetic moment m
magnetic potential A
magnetic susceptibility χm
magnetization curve
magnetization vector M
magnetized domains
moment arm d
orbital and spin magnetic
moments
paramagnetic
solenoid
surface current
density Js
toroid
toroidal coil
torque T
vector Poisson’s
equation
PROBLEMS
B
Section 5-1: Magnetic Forces and Torques
q +
u
a
u
Fm
∗ 5.1 An electron with a speed of 8 × 106 m/s is projected
along the positive x direction into a medium containing
a uniform magnetic flux density B = (x̂4 − ẑ3) T. Given
that e = 1.6 × 10−19 C and the mass of an electron is
me = 9.1 × 10−31 kg, determine the initial acceleration vector
of the electron (at the moment it is projected into the medium).
5.2 When a particle with charge q and mass m is introduced
into a medium with a uniform field B such that the initial
velocity of the particle u is perpendicular to B (Fig. P5.2), the
magnetic force exerted on the particle causes it to move in a
circle of radius a. By equating Fm to the centripetal force on
the particle, determine a in terms of q, m, u, and B.
5.3 The circuit shown in Fig. P5.3 uses two identical springs
to support a 10 cm long horizontal wire with a mass of 20 g.
In the absence of a magnetic field, the weight of the wire
causes the springs to stretch a distance of 0.2 cm each. When a
uniform magnetic field is turned on in the region containing
the horizontal wire, the springs are observed to stretch an
additional 0.5 cm each. What is the intensity of the magnetic
flux density B? The force equation for a spring is F = kd,
where k is the spring constant and d is the distance it has been
stretched.
+
Fm
q
Fm
q +
P
+
q
u
Figure P5.2 Particle of charge q projected with velocity u
into a medium with a uniform field B perpendicular to u
(Problem 5.2).
4Ω
12 V
+ −
Springs
B
10 cm
∗ Answer(s) available in Appendix E.
Figure P5.3 Configuration of Problem 5.3.
266
CHAPTER 5
∗ 5.4 The rectangular loop shown in Fig. P5.4 consists of 20
MAGNETOSTATICS
z
closely wrapped turns and is hinged along the z axis. The plane
of the loop makes an angle of 30◦ with the y axis, and the
current in the windings is 0.5 A. What is the magnitude of
the torque exerted on the loop in the presence of a uniform
field B = ŷ 2.4 T? When viewed from above, is the expected
direction of rotation clockwise or counterclockwise?
20-turn coil
w
l
z
I
I = 0.5 A
y
φ nˆ
20 turns
0.4 m
x
y
30◦
Figure P5.6 Rectangular loop of Problem 5.6.
0.2 m
x
Section 5-2: The Biot–Savart Law
Figure P5.4 Hinged rectangular loop of Problem 5.4.
5.5 In a cylindrical coordinate system, a 2-m long straight
wire carrying a current of 5 A in the positive z direction is
located at r = 4 cm, φ = π /2, and −1 m ≤ z ≤ 1 m.
∗ (a) If B = r̂ 0.2 cos φ (T), what is the magnetic force acting
on the wire?
(b) How much work is required to rotate the wire once about
the z axis in the negative φ direction (while maintaining
r = 4 cm)?
(c) At what angle φ is the force a maximum?
5.6 A 20-turn rectangular coil with sides l = 30 cm and
w = 10 cm is placed in the y–z plane as shown in Fig. P5.6.
(a) If the coil, which carries a current I = 10 A, is in the
presence of a magnetic flux density
B = 2 × 10−2(x̂ + ŷ2)
(T).
Determine the torque acting on the coil.
(b) At what angle φ is the torque zero?
(c) At what angle φ is the torque maximum? Determine its
value.
∗ 5.7 An 8 cm × 12 cm rectangular loop of wire is situated in
the x–y plane with the center of the loop at the origin and its
long sides parallel to the x axis. The loop has a current of 50 A
flowing clockwise (when viewed from above). Determine the
magnetic flux density at the center of the loop.
5.8 Use the approach outlined in Example 5-2 to develop an
expression for the magnetic field H at an arbitrary point P
due to the linear conductor defined by the geometry shown
in Fig. P5.8. If the conductor extends between z1 = 3 m and
z2 = 7 m and carries a current I = 15 A, find H at P = (2, φ , 0).
∗ 5.9 The loop shown in Fig. P5.9 consists of radial lines and
segments of circles whose centers are at point P. Determine
the magnetic field H at P in terms of a, b, θ , and I.
5.10 An infinitely long, thin conducting sheet defined over
the space 0 ≤ x ≤ w and −∞ ≤ y ≤ ∞ is carrying a current
with a uniform surface current density Js = ŷ5 (A/m). Obtain
an expression for the magnetic field at point P = (0, 0, z) in
Cartesian coordinates.
∗ 5.11 An infinitely long wire carrying a 25 A current in the
positive x direction is placed along the x axis in the vicinity of
a 20-turn circular loop located in the x–y plane (Fig. P5.11). If
the magnetic field at the center of the loop is zero, what is the
direction and magnitude of the current flowing in the loop?
PROBLEMS
267
z
P2(z2)
I1 = 6 A
θ2
I2 = 6 A
P
0.5 m
I
θ1
P1(z1)
2m
r
P = (r, φ, z)
Figure P5.12 Arrangement for Problem 5.12.
Figure P5.8 Current-carrying linear conductor of Problem 5.8.
surface. If the magnetic flux density recorded by a magneticfield meter placed at the surface is 15 µ T when the current is
flowing through the cable and 20 µ T when the current is zero,
what is the magnitude of I?
I
b
5.14 Two parallel, circular loops carrying a current of 40 A
each are arranged as shown in Fig. P5.14. The first loop is
situated in the x–y plane with its center at the origin, and the
second loop’s center is at z = 2 m. If the two loops have the
same radius a = 3 m, determine the magnetic field at
(a) z = 0
(b) z = 1 m
(c) z = 2 m
θ
a
P
Figure P5.9 Configuration of Problem 5.9.
1m
z
d=2m
a
z=2m
x
I1
I
Figure P5.11 Circular loop next to a linear current (Problem 5.11).
a
0
5.12 Two infinitely long, parallel wires are carrying 6 A
currents in opposite directions. Determine the magnetic flux
density at point P in Fig. P5.12.
∗ 5.13 A long, East–West-oriented power cable carrying an
unknown current I is at a height of 8 m above the Earth’s
y
I
x
Figure P5.14 Parallel circular loops of Problem 5.14.
268
CHAPTER 5
5.15 A circular loop of radius a carrying current I1 is located
in the x–y plane as shown in Fig. P5.15. In addition, an
infinitely long wire carrying current I2 in a direction parallel
with the z axis is located at y = y0 .
(a) Determine H at P = (0, 0, h).
(b) Evaluate H for a = 3 cm, y0 = 10 cm, h = 4 cm, I1 =
10 A, and I2 = 20 A.
z
MAGNETOSTATICS
5.17 In the arrangement shown in Fig. P5.17, each of the
two long, parallel conductors carries a current I, is supported
by 8-cm long strings, and has a mass per unit length of
1.2 g/cm. Due to the repulsive force acting on the conductors,
the angle θ between the supporting strings is 10◦ . Determine
the magnitude of I and the relative directions of the currents in
the two conductors.
nˆ parallel to zˆ
z
P = (0, 0, h)
I2
F21
θ = 10◦
a
x
y0
y
F12
I1
x
Figure P5.15 Problem 5.15.
y
Figure P5.17 Parallel conductors supported by strings (Problem 5.17).
∗ 5.16 The long, straight conductor shown in Fig. P5.16 lies in
the plane of the rectangular loop at a distance d = 0.1 m. The
loop has dimensions a = 0.2 m and b = 0.5 m, and the currents
are I1 = 20 A and I2 = 30 A. Determine the net magnetic force
acting on the loop.
5.18 An infinitely long, thin conducting sheet of width w
along the x direction lies in the x–y plane and carries a current I
in the −y direction. Determine the following:
∗ (a) The magnetic field at a point P midway between the edges
of the sheet and at a height h above it (Fig. P5.18).
I1
I2
b = 0.5 m
(b) The force per unit length exerted on an infinitely long
wire passing through point P and parallel to the sheet if
the current through the wire is equal in magnitude but
opposite in direction to that carried by the sheet.
5.19 Three long, parallel wires are arranged as shown in
Fig. P5.19. Determine the force per unit length acting on the
wire carrying I3 .
d = 0.1 m
a = 0.2 m
Figure P5.16 Current loop next to a conducting wire (Problem 5.16).
∗ 5.20 A square loop placed as shown in Fig. P5.20 has
2 m sides and carries a current I1 = 5 A. If a straight, long
conductor carrying a current I2 = 10 A is introduced and
placed just above the midpoints of two of the loop’s sides,
determine the net force acting on the loop.
PROBLEMS
269
Section 5-3: Maxwell’s Magnetostatic Equations
I
P
5.21 Current I flows along the positive z direction in the inner
conductor of a long coaxial cable and returns through the outer
conductor. The inner conductor has radius a, and the inner and
outer radii of the outer conductor are b and c, respectively.
I
h
w
Figure P5.18 A linear current source above a current sheet
(Problem 5.18).
(a) Determine the magnetic field in each of the following
regions: 0 ≤ r ≤ a, a ≤ r ≤ b, b ≤ r ≤ c, and r ≥ c.
(b) Plot the magnitude of H as a function of r over the range
from r = 0 to r = 10 cm given that I = 10 A, a = 2 cm,
b = 4 cm, and c = 5 cm.
5.22 A long cylindrical conductor whose axis is coincident
with the z axis has a radius a and carries a current characterized
by a current density J = ẑJ0 /r, where J0 is a constant and
r is the radial distance from the cylinder’s axis. Obtain an
expression for the magnetic field H for
I1 = 10 A
2m
(a) 0 ≤ r ≤ a
(b) r > a
I3 = 10 A
2m
5.23 Repeat Problem 5.22 for a current density J = ẑJ0 e−r .
∗ 5.24 In a certain conducting region, the magnetic field is
2m
given in cylindrical coordinates by
4
H = φ̂φ [1 − (1 + 3r)e−3r].
r
I2 = 10 A
Find the current density J.
Figure P5.19 Three parallel wires of Problem 5.19.
2
H = φ̂φ [1 − (4r + 1)e−4r]
r
z
4
a
a
3
y
I1
2
x
Figure P5.20 Long wire carrying current I2 , just above a
square loop carrying I1 (Problem 5.20).
(A/m)
for r ≤ a,
where a is the conductor’s radius. If a = 5 cm, what is the total
current flowing in the conductor?
I2
1
5.25 A cylindrical conductor whose axis is coincident with
the z axis has an internal magnetic field given by
Section 5-4: Vector Magnetic Potential
5.26 With reference to Fig. 5-10:
∗ (a) Derive an expression for the vector magnetic potential A
at a point P located at a distance r from the wire in the
x–y plane.
(b) Derive B from A. Show that your result is identical with
the expression given by Eq. (5.29), which was derived by
applying the Biot–Savart law.
270
CHAPTER 5
5.27 In a given region of space, the vector magnetic potential
is given by A = x̂5 cos π y + ẑ(2 + sin π x) (Wb/m).
∗ (a) Determine B.
(b) Use Eq. (5.66) to calculate the magnetic flux passing
through a square loop with 0.25-m long edges if the loop
is in the x–y plane, its center is at the origin, and its edges
are parallel to the x and y axes.
Section 5-6: Magnetic Boundary Conditions
5.32 The x–y plane separates two magnetic media with
magnetic permeabilities µ1 and µ2 (Fig. P5.32). If there is
no surface current at the interface and the magnetic field in
medium 1 is
H1 = x̂H1x + ŷH1y + ẑH1z ,
(c) Calculate Φ again using Eq. (5.67).
5.28 A uniform current density given by
J = ẑJ0
(A/m2 )
µ0 J0 2
(x + y2 )
4
find:
(a) H2
(b) θ1 and θ2
(c) Evaluate H2 , θ1 , and θ2 for H1x = 2 (A/m), H1y = 0,
H1z = 4u (A/m), µ1 = µ0 , and µ2 = 4µ0
gives rise to a vector magnetic potential
A = −ẑ
MAGNETOSTATICS
(Wb/m).
z
(a) Apply the vector Poisson’s equation to confirm the above
statement.
(b) Use the expression for A to find H.
(c) Use the expression for J in conjunction with Ampère’s
law to find H. Compare your result with that obtained in
part (b).
∗ 5.29 A thin current element extending between z = −L/2
and z = L/2 carries a current I along +ẑ through a circular
cross-section of radius a.
(a) Find A at a point P located very far from the origin
(assume R is so much larger than L that point P may be
considered to be at approximately the same distance from
every point along the current element).
(b) Determine the corresponding H.
Section 5-5: Magnetic Properties of Materials
5.30 In the model of the hydrogen atom proposed by Bohr
in 1913, the electron moves around the nucleus at a speed of
2 × 106 m/s in a circular orbit of radius 5 × 10−11 m. What
is the magnitude of the magnetic moment generated by the
electron’s motion?
∗ 5.31 Iron contains 8.5 × 1028 atoms/m3 . At saturation, the
alignment of the electrons’ spin magnetic moments in iron can
contribute 1.5 T to the total magnetic flux density B. If the spin
magnetic moment of a single electron is 9.27 × 10−24 (A·m2 ),
how many electrons per atom contribute to the saturated field?
θ1
H1
μ1
x–y plane
μ2
Figure P5.32 Adjacent magnetic media (Problem 5.32).
∗ 5.33 Given that a current sheet with surface current density
Js = x̂ 8 (A/m) exists at y = 0, the interface between two
magnetic media, and H1 = ẑ 11 (A/m) in medium 1 (y > 0),
determine H2 in medium 2 (y < 0).
5.34 In Fig. P5.34, the plane defined by x − y = 1 separates
medium 1 of permeability µ1 from medium 2 of permeability µ2 . If no surface current exists on the boundary and
B1 = x̂2 + ŷ3
(T),
find B2 and then evaluate your result for µ1 = 5µ2 . Hint: Start
by deriving the equation for the unit vector normal to the given
plane.
∗ 5.35 The plane boundary defined by z = 0 separates air from
a block of iron. If B1 = x̂4 − ŷ6 + ẑ8 in air (z ≥ 0), find B2 in
iron (z ≤ 0) given that µ = 5000 µ0 for iron.
PROBLEMS
271
Sections 5-7 and 5-8: Inductance and Magnetic Energy
y
∗ 5.37 Obtain an expression for the self-inductance per unit
Plane x − y = 1
length for the parallel wire transmission line of Fig. 5-28(a)
in terms of a, d, and µ , where a is the radius of the wires,
d is the axis-to-axis distance between the wires, and µ is the
permeability of the medium in which they reside.
x
(1, 0)
Medium 1
μ1
Medium 2
μ2
(0, −1)
Figure P5.34 Magnetic media separated by the plane x −y = 1
(Problem 5.34).
5.38 A solenoid with a length of 20 cm and a radius of 5 cm
consists of 400 turns and carries a current of 12 A. If z = 0
represents the midpoint of the solenoid, generate a plot for
|H(z)| as a function of z along the axis of the solenoid for the
range −20 cm ≤ z ≤ 20 cm in 1 cm steps.
5.39 In terms of the dc current I, how much magnetic energy
is stored in the insulating medium of a 3-m long, air-filled
section of a coaxial transmission line given that the radius of
the inner conductor is 5 cm and the inner radius of the outer
conductor is 10 cm?
∗ 5.40 Determine the mutual inductance between the circular
5.36 Show that, if no surface current densities exist at
the parallel interfaces shown in Fig. P5.36, the relationship
between θ4 and θ1 is independent of µ2 .
loop and the linear current shown in Fig. P5.40.
y
B3
θ4
B2
θ2 θ3
a
x
μ3
d
μ2
I1
B1
μ1
θ1
Figure P5.40 Linear conductor with current I1 next to a
circular loop of radius a at distance d (Problem 5.40).
Figure P5.36 Three magnetic media with parallel interfaces
(Problem 5.36).
Chapter
6
Maxwell's Equations
for Time-Varying Fields
Chapter Contents
6-1
6-2
6-3
6-4
6-5
6-6
6-7
6-8
6-9
6-10
TB12
6-11
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Dynamic Fields, 273
Faraday’s Law, 273
Stationary Loop in a Time-Varying Magnetic
Field, 274
The Ideal Transformer, 278
Moving Conductor in a Static Magnetic
Field, 279
The Electromagnetic Generator, 283
Moving Conductor in a Time-Varying Magnetic
Field, 284
Displacement Current, 285
Boundary Conditions for Electromagnetics, 287
Charge–Current Continuity Relation, 288
Free-Charge Dissipation in a Conductor, 289
EMF Sensors, 290
Electromagnetic Potentials, 292
Chapter 6 Summary, 295
Problems, 296
1. Apply Faraday’s law to compute the voltage induced by a
stationary coil placed in a time-varying magnetic field or
moving in a medium containing a magnetic field.
2. Describe the operation of the electromagnetic generator.
3. Calculate the displacement current associated with a
time-varying electric field.
4. Calculate the rate at which charge dissipates in a material
with known ε and σ .
272
6-1 FARADAY’S LAW
273
Dynamic Fields
Some statements in this and succeeding chapters contradict
conclusions reached in Chapter 4 and 5 as those pertained to
the special case of static charges and dc currents. The behavior
of dynamic fields reduces to that of static ones when ∂ /∂ t is
set to zero.
We begin this chapter by examining Faraday’s and Ampère’s
laws and some of their practical applications. We then combine
Maxwell’s equations to obtain relations among the charge and
current sources, ρv and J; the scalar and vector potentials, V
and A; and the electromagnetic fields, E, D, H, and B. We do
so for the most general time-varying case and for the specific
case of sinusoidal-time variations.
E
lectric charges induce electric fields and electric currents
induce magnetic fields. As long as the charge and current
distributions remain constant in time, so will the fields they
induce. If the charges and currents vary in time, the electric
and magnetic fields vary accordingly. Additionally, however,
the electric and magnetic fields become coupled and travel
through space in the form of electromagnetic waves. Examples
of such waves include light, x-rays, infrared, gamma rays, and
radio waves (see Fig. 1-16).
To study time-varying electromagnetic phenomena, we need
to consider the entire set of Maxwell’s equations simultaneously. These equations, first introduced in the opening section
of Chapter 4, are given in both differential and integral form in
Table 6-1. In the static case (∂ /∂ t = 0), we use the first pair of
Maxwell’s equations to study electric phenomena (Chapter 4)
and the second pair to study magnetic phenomena (Chapter 5).
In the dynamic case (∂ /∂ t 6= 0), the coupling that exists
between the electric and magnetic fields, as expressed by
the second and fourth equations in Table 6-1, prevents such
decomposition. The first equation represents Gauss’s law for
electricity, and it is equally valid for static and dynamic fields.
The same is true for the third equation, which is Gauss’s law
for magnetism. By contrast, the second and fourth equations—
Faraday’s and Ampère’s laws—are of a totally different nature.
Faraday’s law expresses the fact that a time-varying magnetic
field gives rise to an electric field. Conversely, Ampère’s law
states that a time-varying electric field must be accompanied
by a magnetic field.
6-1 Faraday’s Law
The close connection between electricity and magnetism was
established by Oersted, who demonstrated that a wire carrying
an electric current exerts a force on a compass needle and that
φ direction when the curthe needle always turns to point in the φ̂
rent is along the ẑ direction. The force acting on the compass
needle is due to the magnetic field produced by the current
in the wire. Following this discovery, Faraday hypothesized
that if a current produces a magnetic field, then the converse
should also be true: A magnetic field should produce a current
in a wire. To test his hypothesis, he conducted numerous
experiments in his laboratory in London over a period of
about 10 years—all aimed at making magnetic fields induce
currents in wires. Similar work was being carried out by Henry
in Albany, New York. Wires were placed next to permanent
magnets or current-carrying loops of all different sizes, but no
Table 6-1 Maxwell’s equations.
Reference
Gauss’s law
Faraday’s law
No magnetic charges
(Gauss’s law for magnetism)
Ampère’s law
∗ For
a stationary surface S.
Differential Form
Integral Form
Z
∇ · D = ρv
×E = −
∇×
Z
∂B
∂t
C
D · ds = Q
E · dl = −
Z
∇·B = 0
×H = J+
∇×
S
S
∂D
∂t
Z
C
H · dl =
Z
S
(6.1)
∂B
· ds
∂t
B · ds = 0
Z S
J+
∂D
∂t
(6.2)∗
(6.3)
· ds
(6.4)
274
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Loop
Coil
B
I
Galvanometer
I
Battery
Figure 6-1 The galvanometer (predecessor of the ammeter)
shows a deflection whenever the magnetic flux passing through
the square loop changes with time.
opposite direction. Thus, current is induced in the loop when
the magnetic flux changes, and the direction of the current
depends on whether the flux increases (when the battery is
being connected) or decreases (when the battery is being
disconnected). It was further discovered that current can also
flow in the loop while the battery is connected to the coil if
the loop turns or moves either closer to or away from the coil.
The physical movement of the loop changes the amount of flux
linking its surface S, even though the field B due to the coil has
not changed.
A galvanometer is a predecessor of the voltmeter and
ammeter. When a galvanometer detects the flow of current
through the coil, it means that a voltage has been induced
across the galvanometer terminals. This voltage is called the
electromotive force (emf ), Vemf , and the process is called
electromagnetic induction. The emf induced in a closed conducting loop of N turns is given by
Vemf = −N
currents were ever detected. Eventually, these experiments led
to the following discovery by both Faraday and Henry.
◮ Magnetic fields can produce an electric current in a
closed loop, but only if the magnetic flux linking the
surface area of the loop changes with time. The key to
the induction process is change. ◭
To elucidate the induction process, consider the arrangement
shown in Fig. 6-1. A conducting loop connected to a galvanometer (a sensitive instrument used in the 1800s to detect
current flow) is placed next to a conducting coil connected to
a battery. The current in the coil produces a magnetic field B
whose lines pass through the loop. In Section 5-4, we defined
the magnetic flux Φ passing through a loop as the integral of
the normal component of the magnetic flux density over the
surface area of the loop, S, or
Φ=
Z
S
B · ds
dΦ
d
= −N
dt
dt
Z
S
B · ds
(6.5)
Under stationary conditions, the dc current in the coil produces
a constant magnetic field B, which in turn produces a constant
flux through the loop. When the flux is constant, no current
is detected by the galvanometer. However, when the battery
is disconnected, thereby interrupting the flow of current in
the coil, the magnetic field drops to zero, and the consequent
change in magnetic flux causes a momentary deflection of
the galvanometer needle. When the battery is reconnected, the
galvanometer again exhibits a momentary deflection, but in the
(6.6)
Even though the results leading to Eq. (6.6) were also discovered independently by Henry, Eq. (6.6) is attributed to Faraday
and known as Faraday’s law. The significance of the negative
sign in Eq. (6.6) is explained in the next section.
We note that the derivative in Eq. (6.6) is a total time
derivative that operates on the magnetic field B, as well as
the differential surface area ds. Accordingly, an emf can
be generated in a closed conducting loop under any of the
following three conditions:
1. A time-varying magnetic field linking a stationary loop;
tr
.
the induced emf is then called the transformer emf , Vemf
2. A moving loop with a time-varying area (relative to the
normal component of B) in a static field B; the induced
m
.
emf is then called the motional emf , Vemf
3. A moving loop in a time-varying field B.
The total emf is given by
tr
m
Vemf = Vemf
+ Vemf
(Wb).
(V).
m
Vemf
(6.7)
tr
Vemf
= 0 if the loop is stationary [case (1)] and
=0
with
if B is static [case (2)]. For case (3), both terms are important.
Each of the three cases is examined separately in the following
sections.
6-2 Stationary Loop in a Time-Varying
Magnetic Field
The stationary, single-turn, conducting, circular loop with
contour C and surface area S shown in Fig. 6-2(a) is exposed
6-2 STATIONARY LOOP IN A TIME-VARYING MAGNETIC FIELD
Right-Hand Rule for Vemf
Changing B(t)
I
C
1
R
tr
Vemf
Bind
2
S
I
Ri
tr
Vemf
(t)
R
The connection between the direction of ds and the polartr
ity of Vemf
is governed by the following right-hand rule:
If ds points along the thumb of the right hand, then the
direction of the contour C indicated by the four fingers
is such that it always passes across the opening from the
tr
positive terminal of Vemf
to the negative terminal.
If the loop has an internal resistance Ri , the circuit in
Fig. 6-2(a) can be represented by the equivalent circuit shown
in Fig. 6-2(b), in which case the current I flowing through the
circuit is given by
V tr
I = emf .
(6.9)
R + Ri
(a) Loop in a changing B field
1
275
For good conductors, Ri usually is very small, and it may be
ignored in comparison with practical values of R.
2
(b) Equivalent circuit
Figure 6-2 (a) Stationary circular loop in a changing magnetic
field B(t) and (b) its equivalent circuit.
to a time-varying magnetic field B(t). As stated earlier, the
emf induced when S is stationary and the field is time varying
tr
. Since the
is called the transformer emf and is denoted Vemf
loop is stationary, d/dt in Eq. (6.6) now operates on B(t) only.
Hence,
tr
Vemf
= −N
Z
∂B
· ds,
S ∂t
(transformer emf)
(6.8)
where the full derivative d/dt has been moved inside the
integral and changed into a partial derivative ∂ /∂ t to signify
that it operates on B only. The transformer emf is the voltage
difference that would appear across the small opening between
terminals 1 and 2, even in the absence of the resistor R. That
tr
is, Vemf
= V12 , where V12 is the open-circuit voltage across the
tr
open ends of the loop. Under dc conditions, Vemf
= 0. For
the loop shown in Fig. 6-2(a) and the associated definition
tr
for Vemf
given by Eq. (6.8), the direction of ds, which is the
loop’s differential surface normal, can be chosen to be either
upward or downward. The two choices are associated with
opposite designations of the polarities of terminals 1 and 2 in
Fig. 6-2(a).
tr
◮ The polarity of Vemf
and hence the direction of I is
governed by Lenz’s law, which states that the current in
the loop is always in a direction that opposes the change
of magnetic flux Φ(t) that produced I. ◭
The current I induces a magnetic field of its own, Bind , with a
corresponding flux Φind . The direction of Bind is governed by
the right-hand rule: If I is in a clockwise direction, then Bind
points downward through S; conversely, if I is in a counterclockwise direction, then Bind points upward through S. If the
original field B(t) is increasing, which means that dΦ/dt > 0,
then according to Lenz’s law I has to be in the direction shown
in Fig. 6-2(a) in order for Bind to be in opposition to B(t).
Consequently, terminal 2 would be at a higher potential than
tr
terminal 1, and Vemf
would have a negative value. However, if
B(t) were to remain in the same direction but to decrease in
magnitude, then dΦ/dt would become negative, the current
would have to reverse direction, and its induced field Bind
would be in the same direction as B(t) in order to oppose the
tr
change (decrease) of B(t). In that case, Vemf
would be positive.
◮ It is important to remember that Bind serves to oppose
the change in B(t) and not necessarily B(t) itself. ◭
Despite the presence of the small opening between terminals
1 and 2 of the loop in Fig. 6-2(a), we shall treat the loop as a
closed path with contour C. We do this in order to establish
the link between B and the electric field E associated with
tr
the induced emf, Vemf
. Also, at any point along the loop, the
276
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
field E is related to the current I flowing through the loop. For
tr
contour C, Vemf
is related to E by
tr
Vemf
=
Z
C
z
B
E · dl.
Z
S
× E) · ds = −
(∇×
Z
∂B
· ds,
S ∂t
(6.12)
and in order for the two integrals to be equal for all possible
choices of S, their integrands must be equal, which gives
×E = −
∇×
∂B
.
∂t
(Faraday’s law)
(6.13)
This differential form of Faraday’s law states that a timevarying magnetic field induces an electric field E whose curl
is equal to the negative of the time derivative of B. Even
though the derivation leading to Faraday’s law started out
by considering the field associated with a physical circuit,
Eq. (6.13) applies at any point in space, whether or not a
physical circuit exists at that point.
Example 6-1:
1
R
For N = 1 (a loop with one turn), equating Eqs. (6.8) and (6.10)
gives
Z
Z
∂B
· ds,
(6.11)
E · dl = −
C
S ∂t
which is the integral form of Faraday’s law given in Table 6-1.
We should keep in mind that the direction of the contour C and
the direction of ds are related by the right-hand rule.
By applying Stokes’s theorem to the left-hand side of
Eq. (6.11), we have
B
I
(6.10)
tr
Vemf
y
a
2
N turns
Figure 6-3 Circular loop with N turns in the x–y plane. The
magnetic field is B = B0 (ŷ2 + ẑ3) sin ω t (Example 6-1).
Solution: (a) With ds defined upwards (along +ẑ direction),
the magnetic flux linking each turn of the inductor is
Φ=
Z
S
B · ds =
Z
S
[B0 (ŷ 2 + ẑ3) sin ω t] · ẑ ds = 3π a2B0 sin ω t.
tr
(b) To find Vemf
, we can apply Eq. (6.8) or apply the general
expression given by Eq. (6.6) directly. The latter approach
gives
tr
Vemf
= −N
d
dΦ
= − (3π Na2 B0 sin ω t)
dt
dt
= −3π N ω a2B0 cos ω t.
Having chosen the direction of ds to be upwards, the polarity
tr
of Vemf
is governed by the right-hand rule requiring the direction of the contour C in Fig. 6-3 to pass across the gap from
tr
the positive terminal of Vemf
to the negative terminal. That is,
Inductor in a Changing
Magnetic Field
tr
Vemf
= V1 − V2 = −3π N ω a2B0 cos ω t.
For N = 10, a = 0.1 m, ω = 103 rad/s, and B0 = 0.2 T,
An inductor is formed by winding N turns of a thin conducting
wire into a circular loop of radius a. The inductor loop is in
the x–y plane with its center at the origin and connected to a
resistor R, as shown in Fig. 6-3. In the presence of a magnetic
field B = B0 (ŷ2 + ẑ3) sin ω t, where ω is the angular frequency,
find
(a) the magnetic flux linking a single turn of the inductor,
(b) the transformer emf given that N = 10, B0 = 0.2 T,
a = 10 cm, and ω = 103 rad/s,
tr
(c) the polarity of Vemf
at t = 0, and
(d) the induced current in the circuit for R = 1 kΩ (assume
the wire resistance to be much smaller than R).
tr
Vemf
= −188.5 cos103t
(V).
(c) The flux passing through each loop in the upward direction
is Φ(t) = 3π a2 B0 sin ω t. At t = 0, Φ(0) = 0, but
dΦ
dt
= 3π a2 ω B0 cos ω t0
t=0
t=0
= 3π a2ω B0 > 0.
Hence, even though the flux itself is zero at t = 0, its time
derivative is positive. Therefore, the flux is increasing at t = 0.
Lenz’s law requires the direction of the current I to be such
that the flux it induces, Φind , opposes the change in Φ. Since
Φ is defined to be upwards, and it is increasing at t = 0,
6-2 STATIONARY LOOP IN A TIME-VARYING MAGNETIC FIELD
277
Module 6.1 Circular Loop in Time-Varying Magnetic Field Faraday’s law of induction is demonstrated by simulating
the current induced in a loop in response to the change in magnetic flux flowing through it.
Φind should be downward through the loop in order to slow
down the increase in Φ. Hence, I should be in the direction
shown in Fig. 6-3.
Since I flows through a resistor from its positive voltage
terminal to its negative voltage terminal, it follows that V12
should be negative at t = 0, which it is:
tr
Vemf
= V1 − V2 = −188.5
replaced with a 10-turn square loop centered at the origin
and having 20 cm sides oriented parallel to the x and
y axes. If B = ẑB0 x2 cos 103t and B0 = 100 T, find the
current in the circuit.
Answer: I = −133 sin 103t (mA). (See
(V).
(d) The current I is given by
I=
Exercise 6-2: Suppose that the loop of Example 6-1 is
.)
Example 6-2: Lenz’s Law
V2 − V1 188.5
cos 103t = 0.19 cos103t
=
R
103
(A).
tr
Exercise 6-1: For the loop shown in Fig. 6-3, what is Vemf
if B = ŷB0 cos ω t?
Determine voltages V1 and V2 across the 2 Ω and 4 Ω resistors
shown in Fig. 6-4. The loop is located in the x–y plane, its area
is 4 m2 , the magnetic flux density is B = −ẑ0.3t (T), and the
internal resistance of the wire may be ignored.
Solution: The flux flowing through the loop is
tr
Answer: Vemf
= 0 because B is orthogonal to the loop’s
surface normal ds. (See
EM
EM
.)
Φ=
Z
S
B · ds =
Z
S
(−ẑ0.3t) · ẑ ds = −0.3t × 4 = −1.2t
(Wb),
278
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
6-3 The Ideal Transformer
I
y
x
4Ω
V2
V1
2Ω
Area = 4 m2
B
Figure 6-4 Circuit for Example 6-2.
and the corresponding transformer emf is
tr
Vemf
=−
dΦ
= 1.2
dt
(V).
Since the magnetic flux through the loop is along the
−z direction (into the page) and increases in magnitude with
time t, Lenz’s law states that the induced current I should be in
a direction such that the magnetic flux density Bind it induces
counteracts the direction of change of Φ. Hence, I has to be
in the direction shown in the circuit because the corresponding
Bind is along the +z direction in the region inside the loop area.
This, in turn, means that V1 and V2 are positive voltages.
The total voltage of 1.2 V is distributed across two resistors
in series. Consequently,
I=
The transformer shown in Fig. 6-5(a) consists of two coils
wound around a common magnetic core. The primary coil has
N1 turns and is connected to an ac voltage source V1 (t). The
secondary coil has N2 turns and is connected to a load resistor
RL . In an ideal transformer the core has infinite permeability
(µ = ∞), and the magnetic flux is confined within the core.
◮ The directions of the currents flowing in the two
coils, I1 and I2 , are defined such that, when I1 and I2
are both positive, the flux generated by I2 is opposite
to that generated by I1 . The transformer gets its name
from the fact that it transforms currents, voltages, and
impedances between its primary and secondary circuits,
and vice versa. ◭
On the primary side of the transformer, the voltage source
V1 generates current I1 in the primary coil, which establishes
a flux Φ in the magnetic core. The flux Φ and voltage V1 are
I1
Ф
V1(t)
N1
I2
N2
V2(t) RL
Ф
tr
Vemf
1.2
=
= 0.2 A,
R1 + R2 2 + 4
(a)
Magnetic core
and
V1 = IR1 = 0.2 × 2 = 0.4 V,
V2 = IR2 = 0.2 × 4 = 0.8 V.
I1
Ф
V1(t)
N1
I2
N2
Concept Question 6-1:
V2(t) RL
Explain Faraday’s law and the
function of Lenz’s law.
Ф
Under what circumstances is
the net voltage around a closed loop equal to zero?
Concept Question 6-2:
Suppose the magnetic flux density linking the loop of Fig. 6-4 (Example 6-2) is given
by B = −ẑ0.3e−t (T). What would the direction of the
current be, relative to that shown in Fig. 6-4, for t ≥ 0?
Concept Question 6-3:
(b)
Figure 6-5 In a transformer, the directions of I1 and I2 are
such that the flux Φ generated by one of them is opposite to that
generated by the other. The direction of the secondary winding
in (b) is opposite to that in (a), and so are the direction of I2 and
the polarity of V2 .
6-4 MOVING CONDUCTOR IN A STATIC MAGNETIC FIELD
279
related by Faraday’s law:
dΦ
V1 = −N1
.
dt
A similar relation holds true on the secondary side:
dΦ
.
V2 = −N2
dt
The combination of Eqs. (6.14) and (6.15) gives
I1(t)
(6.14)
(6.15)
Figure 6-6 Equivalent circuit for the primary side of the
transformer.
V1 N1
=
.
V2 N2
(6.16)
In an ideal lossless transformer, all the instantaneous power
supplied by the source connected to the primary coil is delivered to the load on the secondary side. Thus, no power is lost
in the core, and
P1 = P2 .
(6.17)
Since P1 = I1V1 and P2 = I2V2 , and in view of Eq. (6.16), we
have
N2
I1
=
.
I2
N1
Zin =
2
× B).
Fm = q(u×
(6.22)
This magnetic force is equivalent to the electrical force that
would be exerted on the particle by the electric field Em
given by
Fm
× B.
= u×
(6.23)
Em =
q
The field Em generated by the motion of the charged particle
is called a motional electric field and is in the direction
perpendicular to the plane containing u and B. For the wire
shown in Fig. 6-7, Em is along −ŷ. The magnetic force acting
B
B
y
(6.19)
Em
ZL .
u
l
z
(6.21)
x
u
(6.20)
When the load is an impedance ZL and V1 is a sinusoidal
source, the phasor-domain equivalent of Eq. (6.20) is
N1
N2
Consider a wire of length l moving across a static magnetic
field B = ẑB0 with constant velocity u, as shown in Fig. 6-7.
The conducting wire contains free electrons. From Eq. (5.3),
the magnetic force Fm acting on a particle with charge q
moving with velocity u in a magnetic field B is
1
V1
.
I1
Use of Eqs. (6.16) and (6.18) gives
2
V2 N1 2
N1
Rin =
=
RL .
I2 N2
N2
6-4 Moving Conductor in a Static
Magnetic Field
(6.18)
Thus, whereas the ratio of the voltages given by Eq. (6.16) is
proportional to the corresponding turns ratio, the ratio of the
currents is equal to the inverse of the turns ratio. If N1 /N2 =
0.1, V2 of the secondary circuit would be 10 times V1 of the
primary circuit, but I2 would be only I1 /10.
The transformer shown in Fig. 6-5(b) is identical to that
in Fig. 6-5(a) except for the direction of the windings of the
secondary coil. Because of this change, the direction of I2 and
the polarity of V2 in Fig. 6-5(b) are the reverse of those in
Fig. 6-5(a).
The voltage and current in the secondary circuit in
Fig. 6-5(a) are related by V2 = I2 RL . To the input circuit,
the transformer may be represented by an equivalent input
resistance Rin , as shown in Fig. 6-6, defined as
Rin =
Rin
V1(t)
2
Moving
wire
Magnetic field line
(out of the page)
Figure 6-7 Conducting wire moving with velocity u in a static
magnetic field.
280
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Note that B increases linearly with x. The bar starts from x = 0
at t = 0. Find the motional emf between terminals 1 and 2 and
the current I flowing through the resistor R. Assume that the
loop resistance Ri ≪ R.
on the (negatively charged) electrons in the wire causes them
to drift in the direction of −Em ; that is, toward the wire end
labeled 1 in Fig. 6-7. This, in turn, induces a voltage difference
between ends 1 and 2, with end 2 being at the higher potential.
m
The induced voltage is called a motional emf , Vemf
, and is
defined as the line integral of Em between ends 2 and 1 of the
wire,
m
Vemf
= V12 =
Z 1
2
Em · dl =
Z 1
2
× B) · dl.
(u×
Solution: This problem can be solved by using the motional
emf expression given by Eq. (6.26) or by applying the general formula of Faraday’s law. We now show that the two
approaches yield the same result.
(6.24)
(a) Motional emf
For the conducting wire, u × B = x̂u × ẑB0 = −ŷuB0 and
dl = ŷ dl. Hence,
m
Vemf
= V12 = −uB0 l.
The sliding bar, being the only part of the circuit that crosses
the lines of the field B, is the only part of contour 2341 that
m
contributes to Vemf
. Hence, at x = x0 , for example,
(6.25)
In general, if any segment of a closed circuit with contour C
moves with a velocity u across a static magnetic field B, then
the induced motional emf is given by
m
Vemf
=
Z
C
m
Vemf
= V12 = V43 =
=
Z 4
3
Z 4
3
× B) · dl.
(u×
(motional emf)
(6.26)
m
Vemf
= −B0 u2 lt
(6.27)
tr
m
Since B is static, Vemf
= 0 and Vemf = Vemf
only. To verify
that the same result can be obtained by the general form of
Faraday’s law, we evaluate the flux Φ through the surface of
the loop. Thus,
The rectangular loop shown in Fig. 6-8 has a constant width l,
but its length x0 increases with time as a conducting bar slides
with uniform velocity u in a static magnetic field B = ẑB0 x.
Φ=
Z
S
B · ds =
Z
S
(ẑB0 x) · ẑ dx dy = B0 l
4
Z x0
x dx =
0
y
dl
I 1
u
l
(V).
(b) Total emf
Sliding Bar
R
(x̂u × ẑB0 x0 ) · ŷ dl = −uB0x0 l.
The length of the loop is related to u by x0 = ut. Hence,
◮ Only those segments of the circuit that cross magnetic
m
field lines contribute to Vemf
.◭
Example 6-3:
× B) · dl
(u×
m
Vemf
2
u
z
x
Magnetic field B
3
x=0
x0
Figure 6-8 Sliding bar with velocity u in a magnetic field that increases linearly with x; that is, B = ẑB0 x (Example 6-3).
B0 lx20
.
2
(6.28)
6-4 MOVING CONDUCTOR IN A STATIC MAGNETIC FIELD
Substituting x0 = ut in Eq. (6.28) and then evaluating the
negative of the derivative of the flux with respect to time gives
Vemf = −
dΦ
d
=−
dt
dt
B0 lu2t
2
2
281
The induced voltage V12 is then given by
V12 =
= −B0 u2 lt
(V),
(6.29)
=
Z 1
2
× B(y1 )] · dl
[u×
Z −l/2
l/2
which is identical with Eq. (6.27). Since V12 is negative, the
current I = B0 u2 lt/R flows in the direction shown in Fig. 6-8.
(ŷ5 × ẑ0.2e−0.2) · x̂ dx
= −e−0.2 l
= −2e−0.2
Example 6-4:
= −1.637
Moving Loop
(V).
Similarly,
The rectangular loop shown in Fig. 6-9 is situated in the
x–y plane and moves away from the origin with velocity
u = ŷ5 (m/s) in a magnetic field given by
B(y) = ẑ 0.2e−0.1y
V43 = −u B(y2 ) l
= −5 × 0.2e−0.25 × 2
= −1.558
(T).
If R = 5 Ω, find the current I at the instant that the loop sides
are at y1 = 2 m and y2 = 2.5 m. The loop resistance may be
ignored.
× B is along x̂, voltages are induced across
Solution: Since u×
only the sides oriented along x̂, namely the sides linking point
1 to 2 and point 3 to 4. Had B been uniform, the induced
voltages would have been the same, and the net voltage across
the resistor would have been zero. In the present case, however,
B decreases exponentially with y, thereby assuming a different
value over side 1–2 than over side 3–4. Side 1–2 is at y1 = 2 m,
and the corresponding magnetic field is
B(y1 ) = ẑ 0.2e−0.1y1 = ẑ 0.2e−0.2
(V).
Consequently, the current is in the direction shown in the figure
and its magnitude is
I=
V43 − V12 0.079
=
= 15.8 (mA).
R
5
Example 6-5: Moving Rod Next to a Wire
The wire shown in Fig. 6-10 carries a current I = 10 A.
A 30-cm long metal rod moves with a constant velocity u = ẑ5
m/s. Find V12 , where 1 and 2 are the two ends of the rod.
(T).
z
z
1
I
4
B
u
l=2m
V12
y1 = 2 m
V43
y2 = 2.5 m
I = 10 A
r
y
2
3
R
0.5 m
Figure 6-9 Moving loop of Example 6-4.
u
B
1
2
10 cm
B
Metal rod
B
Wire
u
x
B
30 cm
B
Figure 6-10 Moving rod of Example 6-5.
282
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Module 6.2 Rotating Wire Loop in Constant Magnetic Field The principle of the electromagnetic generator is
demonstrated by a rectangular loop rotating in the presence of a magnetic field.
Solution: The current I induces a magnetic field
µ0 I
B = φ̂φ
,
2π r
where r is the radial distance from the wire and the direction
of φ̂φ is into the page on the rod side of the wire. The movement
of the rod in the presence of the field B induces a motional emf
given by
V12 =
=
Z 10 cm
× B) · dl
(u×
40 cm
Z 10 cm 40 cm
=−
=−
5 µ0 I
2π
µ0 I
ẑ 5 × φ̂φ
· r̂ dr
2π r
Z 10 cm
dr
40 cm
r
5 × 4π × 10−7 × 10
= 13.9
2π
(µ V).
10
× ln
40
Suppose that no friction is
involved in sliding the conducting bar of Fig. 6-8 and
that the horizontal arms of the circuit are very long.
Hence, if the bar is given an initial push, it should
continue moving at a constant velocity, and its movement
generates electrical energy in the form of an induced emf,
indefinitely. Is this a valid argument? If not, why not? Can
we generate electrical energy without having to supply an
equal amount of energy by other means?
Concept Question 6-4:
Is the current flowing in the rod
of Fig. 6-10 a steady current? Examine the force on a
charge q at ends 1 and 2 and compare.
Concept Question 6-5:
Exercise 6-3: For the moving loop of Fig. 6-9, find I when
the loop sides are at y1 = 4 m and y2 = 4.5 m. Also, reverse
the direction of motion such that u = −ŷ5 (m/s).
Answer: I = −13 (mA). (See
EM
.)
6-5 THE ELECTROMAGNETIC GENERATOR
283
Exercise 6-4: Suppose that we turn the loop of Fig. 6-9
so that its surface is parallel to the x–z plane. What would
I be in that case?
Answer: I = 0. (See
EM
.)
S
2
B
1
6-5 The Electromagnetic Generator
R
The electromagnetic generator is the converse of the electromagnetic motor. The principles of operation of both instruments may be explained with the help of Fig. 6-11. A permanent magnet is used to produce a static magnetic field B in the
slot between its two poles. When a current is passed through
the conducting loop, as depicted in Fig. 6-11(a), the current
flows in opposite directions in segments 1–2 and 3–4 of the
loop. The induced magnetic forces on the two segments are
also opposite, resulting in a torque that causes the loop to rotate
about its axis. Thus, in a motor, electrical energy supplied by a
voltage source is converted into mechanical energy in the form
of a rotating loop, which can be coupled to pulleys, gears, or
other movable objects.
If instead of passing a current through the loop to make it
turn the loop is made to rotate by an external force, the movement of the loop in the magnetic field produces a motional emf,
m
Vemf
, as shown in Fig. 6-11(b). Hence, the motor has become
a generator, and mechanical energy is getting converted into
electrical energy.
Let us examine the operation of the electromagnetic generator in more detail using the coordinate system shown in
Fig. 6-12. The magnetic field is
B = ẑB0 ,
(6.30)
and the axis of rotation of the conducting loop is along the x
axis. Segments 1–2 and 3–4 of the loop are of length l each,
and both cross the magnetic flux lines as the loop rotates. The
other two segments are each of width w, and neither crosses
the B lines when the loop rotates. Hence, only segments 1–2
m
and 3–4 contribute to the generation of the motional emf, Vemf
.
As the loop rotates with an angular velocity ω about its own
axis, segment 1–2 moves with velocity u given by
u = n̂ω
w
,
2
3
I
(6.31)
where n̂, the surface normal to the loop, makes an angle α with
the z axis. Hence,
× ẑ = x̂ sin α .
(6.32)
n̂×
Segment 3–4 moves with velocity −u. Application of
I 4
V(t)
N
Magnet
ω
Axis of rotation
(a) ac motor
S
2
B
1
3
I
m
Vemf
R
I 4
N
Magnet
ω
Axis of rotation
(b) ac generator
Figure 6-11 Principles of the ac motor and the ac generator.
In (a) the magnetic torque on the wires causes the loop to rotate
and in (b) the rotating loop generates an emf.
Eq. (6.26), consistent with our choice of n̂, gives
m
Vemf
= V14 =
=
Z 1
2
× B) · dl +
(u×
Z l/2 h
Z 3
4
× B) · dl
(u×
i
w
× ẑB0 · x̂ dx
2
−l/2
Z −l/2 h
i
w
× ẑB0 · x̂ dx.
−n̂ω
+
2
l/2
n̂ω
(6.33)
Using Eq. (6.32) in Eq. (6.33), we obtain the result
m
Vemf
= wl ω B0 sin α = Aω B0 sin α ,
(6.34)
284
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Contrast the operation of an ac
motor with that of an ac generator.
Concept Question 6-6:
z
The rotating loop of Fig. 6-12
had a single turn. What would be the emf generated by a
loop with 10 turns?
Concept Question 6-7:
2
B
l
w
dl
α
nˆ
1
3
y
Slip
rings
m
Vemf
4
Loop surface
normal
6-6 Moving Conductor in a
Time-Varying Magnetic Field
ω
Brushes
x
Figure 6-12 A loop rotating in a magnetic field induces an
emf.
where A = wl is the surface area of the loop. The angle α is
related to ω by
α = ω t + C0 ,
(6.35)
where C0 is a constant determined by initial conditions. For
example, if α = 0 at t = 0, then C0 = 0. In general,
m
Vemf
= Aω B0 sin(ω t + C0 )
(V).
(6.36)
This same result can also be obtained by applying the
general form of Faraday’s law given by Eq. (6.6). The flux
linking the surface of the loop is
Φ=
Z
S
B · ds =
Z
S
The magnetic flux linking the
loop shown in Fig. 6-12 is maximum when α = 0 (loop in
x–y plane), yet according to Eq. (6.34), the induced emf
is zero when α = 0. Conversely, when α = 90◦ , the flux
m
linking the loop is zero, but Vemf
is at a maximum. Is this
consistent with your expectations? Why?
Concept Question 6-8:
ẑB0 · n̂ ds = B0 A cos α = B0 A cos(ω t + C0 ),
(6.37)
and
dΦ
d
= − [B0 A cos(ω t +C0 )] = Aω B0 sin(ω t +C0 ),
dt
dt
(6.38)
which is identical with the result given by Eq. (6.36).
For the general case of a single-turn conducting loop moving
in a time-varying magnetic field, the induced emf is the sum
of a transformer component and a motional component. Thus,
the sum of Eqs. (6.8) and (6.26) gives
tr
m
Vemf = Vemf
+ Vemf
=−
Z
∂B
· ds +
S ∂t
Z
C
(u × B) · dl.
(6.39)
Vemf is also given by the general expression of Faraday’s law:
Vemf = −
d
dΦ
=−
dt
dt
Z
S
B · ds (total emf).
(6.40)
In fact, it can be shown mathematically that the right-hand side
of Eq. (6.39) is equivalent to the right-hand side of Eq. (6.40).
For a particular problem, the choice between using Eq. (6.39)
or Eq. (6.40) is usually made on the basis of which is the easier
to apply. In either case, for an N-turn loop, the right-hand sides
of Eqs. (6.39) and (6.40) should be multiplied by N.
Vemf = −
◮ The voltage induced by the rotating loop is sinusoidal
in time with an angular frequency ω equal to that of the
rotating loop, and its amplitude is equal to the product
of the surface area of the loop, the magnitude of the
magnetic field generated by the magnet, and the angular
frequency ω . ◭
Example 6-6: Electromagnetic Generator
Find the induced voltage when the rotating loop of the electromagnetic generator of Section 6-5 is in a magnetic field
B = ẑB0 cos ω t. Assume that α = 0 at t = 0.
Solution: The flux Φ is given by Eq. (6.37) with B0 replaced
with B0 cos ω t. Thus,
Φ = B0 A cos2 ω t,
6-7 DISPLACEMENT CURRENT
285
and
The second term on the right-hand side of Eq. (6.43) of course
has the same unit (amperes) as the current Ic , and because
it is proportional to the time derivative of the electric flux
density D, which is also called the electric displacement, it is
called the displacement current Id . That is,
dΦ
Vemf = −
dt
d
= − (B0 A cos2 ω t)
dt
= 2B0 Aω cos ω t sin ω t = B0 Aω sin 2ω t.
Id =
6-7 Displacement Current
∂D
×H = J+
∇×
(Ampère’s law).
(6.41)
∂t
Integrating both sides of Eq. (6.41) over an arbitrary open
surface S with contour C, we have
Z
Z
Z
∂D
× H) · ds = J · ds +
· ds.
(6.42)
(∇×
S ∂t
S
S
The surface integral of J equals the conduction current Ic
flowing through S, and the surface integral of ∇ × H can be
converted into a line integral of H over the contour C bounding
C by invoking Stokes’s theorem. That is,
and
Z
S
Z
S
J · ds = Ic ,
Z
× H) · ds =
(∇×
C
H · dl = Ic +
C
H · dl,
Z
∂D
· ds.
S ∂t
S
Jd · ds =
Z
∂D
· ds,
S ∂t
(6.44)
where Jd = ∂ D/∂ t represents a displacement current density.
In view of Eq. (6.44),
Ampère’s law in differential form is given by
Z
Z
(6.43)
(Ampère’s law)
Z
C
H · dl = Ic + Id = I,
(6.45)
where I is the total current. In electrostatics, ∂ D/∂ t = 0; therefore, Id = 0 and I = Ic . The concept of displacement current
was first introduced in 1873 by James Clerk Maxwell when
he formulated the unified theory of electricity and magnetism
under time-varying conditions.
The parallel-plate capacitor is commonly used as an example with which to illustrate the physical meaning of the
displacement current Id . The simple circuit shown in Fig. 6-13
consists of a capacitor and an ac source with voltage Vs (t)
given by
(V).
(6.46)
Vs (t) = V0 cos ω t
According to Eq. (6.45), the total current flowing through
any surface consists, in general, of a conduction current Ic
and a displacement current Id . Let us find Ic and Id through
each of the following two imaginary surfaces: (1) the cross
section of the conducting wire, S1 , and (2) the cross section
of the capacitor S2 (Fig. 6-13). We denote the conduction
and displacement currents in the wire as I1c and I1d and those
through the capacitor as I2c and I2d .
I1= I1c
Imaginary
surface S1
Vs(t)
+ + + + + + +
E
I2d
Imaginary
surface S2
– – – – – – –
y
Figure 6-13 The displacement current I2d in the insulating material of the capacitor is equal to the conducting current I1c in the wire.
286
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
In the perfectly conducting wire, D = E = 0; hence,
Eq. (6.44) gives I1d = 0. As for I1c , we know from circuit
theory that it is related to the voltage across the capacitor VC
by
dVC
dt
d
= C (V0 cos ω t)
dt
= −CV0 ω sin ω t,
I1c = C
(6.47)
where we used the fact that VC = Vs (t). With I1d = 0, the total
current in the wire is simply I1 = I1c = −CV0 ω sin ω t.
In the perfect dielectric with permittivity ε between the
capacitor plates, σ = 0. Hence, I2c = 0 because no conduction
current exists there. To determine I2d , we need to apply
Eq. (6.44). From Example 4-14, the electric field E in the
dielectric spacing is related to the voltage Vc across its plates
by
Vc
V0
E = ŷ
(6.48)
= ŷ cos ω t,
d
d
where d is the spacing between the plates and ŷ is the direction
from the higher-potential plate toward the lower-potential plate
at t = 0. The displacement current I2d is obtained by applying
Eq. (6.44) with ds = ŷ ds:
Z
S
Displacement Current
Density
The conduction current flowing through a wire with conductivity σ = 2 × 107 S/m and relative permittivity εr = 1 is given
by Ic = 2 sin ω t (mA). If ω = 109 rad/s, find the displacement
current.
Solution: The conduction current Ic = JA = σ EA, where A is
the cross section of the wire. Hence,
E=
2 × 10−3 sin ω t
Ic
=
σA
2 × 107A
1 × 10−10
sin ω t
(V/m).
A
Application of Eq. (6.44), with D = ε E, leads to
Id = Jd A
∂E
∂t
∂ 1 × 10−10
= εA
sin ω t
∂t
A
= εA
εA
= − V0 ω sin ω t
d
= −CV0 ω sin ω t,
Example 6-7:
=
∂D
· ds
∂t
Z ∂
ε V0
ŷ
=
cos ω t ·(ŷ ds)
d
A ∂t
I2d =
has a finite conductivity σw , then D in the wire would not be
zero; therefore, the current I1 would consist of a conduction
current I1c as well as a displacement current I1d ; that is,
I1 = I1c + I1d . By the same token, if the dielectric spacing
material has a nonzero conductivity σd , then free charges
would flow between the two plates, and I2c would not be zero.
In that case, the total current flowing through the capacitor
would be I2 = I2c + I2d . No matter the circumstances, the total
capacitor current remains equal to the total current in the wire.
That is, I1 = I2 .
(6.49)
where we used the relation C = ε A/d for the capacitance of
the parallel-plate capacitor with plate area A. The expression
for I2d in the dielectric region between the conducting plates
is identical with that given by Eq. (6.47) for the conduction
current I1c in the wire. The fact that these two currents are
equal ensures the continuity of total current flow through the
circuit.
◮ Even though the displacement current does not transport free charges, it nonetheless behaves like a real
current. ◭
In the capacitor example, we treated the wire as a perfect
conductor, and we assumed that the space between the capacitor plates was filled with a perfect dielectric. If the wire
= εω × 10−10 cos ω t
= 0.885 × 10−12 cos ω t
(A),
where we used ω = 109 rad/s and ε = ε0 = 8.85 × 10−12 F/m.
Note that Ic and Id are in phase quadrature (90◦ phase shift
between them). Also, Id is about nine orders of magnitude
smaller than Ic , which is why the displacement current usually
is ignored in good conductors.
Exercise 6-5: A poor conductor is characterized by a
conductivity σ = 100 (S/m) and permittivity ε = 4ε0 .
At what angular frequency ω is the amplitude of the
conduction current density J equal to the amplitude of the
displacement current density Jd ?
Answer: ω = 2.82 × 1012 (rad/s). (See
EM
.)
6-8 BOUNDARY CONDITIONS FOR ELECTROMAGNETICS
287
Module 6.3 Displacement Current Observe the displacement current through a parallel-plate capacitor.
6-8 Boundary Conditions for
Electromagnetics
In Chapters 4 and 5, we applied the integral form of Maxwell’s
equations under static conditions to obtain boundary conditions applicable to the tangential and normal components
of E, D, B, and H on interfaces between contiguous media
(Section 4-8 for E and D and in Section 5-6 for B and H). In the
dynamic case, Maxwell’s equations (Table 6-1) include two
new terms not accounted for in electrostatics and magnetostatics, namely, ∂ B/∂ t in Faraday’s law and ∂ D/∂ t in Ampère’s
law.
◮ Nevertheless, the boundary conditions derived previously for electrostatic and magnetostatic fields remain
valid for time-varying fields as well. ◭
This is because, if we were to apply the procedures outlined in
the previously referenced sections for time-varying fields, we
would find that the combination of the aforementioned terms
vanish as the areas of the rectangular loops in Figs. 4-20 and
5-24 are made to approach zero.
The combined set of electromagnetic boundary conditions
is summarized in Table 6-2.
When conduction current flows
through a material, a certain number of charges enter the
material on one end and an equal number leave on the
other end. What’s the situation like for the displacement
current through a perfect dielectric?
Concept Question 6-9:
Verify that the integral form
of Ampère’s law given by Eq. (6.43) leads to the boundary
condition that the tangential component of H is continuous across the boundary between two dielectric media.
Concept Question 6-10:
288
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Table 6-2 Boundary conditions for the electric and magnetic fields.
Field Components
General Form
Medium 1
Dielectric
Medium 2
Dielectric
Medium 1
Dielectric
Medium 2
Conductor
Tangential E
E1t = E2t
E1t = E2t = 0
n̂2 × (E1 − E2 ) = 0
n̂2 ·(D1 − D2 ) = ρs
Normal D
D1n − D2n = ρs
D1n = ρs
D2n = 0
Tangential H
H1t = H2t
H1t = Js
H2t = 0
n̂2 × (H1 − H2 ) = Js
Normal B
B1n = B2n
B1n = B2n = 0
n̂2 · (B1 − B2 ) = 0
Notes: (1) ρs is the surface charge density at the boundary; (2) Js is the surface current density at the boundary; (3) normal
components of all fields are along n̂2 , the outward unit vector of medium 2; (4) E1t = E2t implies that the tangential
components are equal in magnitude and parallel in direction; (5) direction of Js is orthogonal to (H1 − H2 ).
6-9 Charge–Current Continuity Relation
Under static conditions, the charge density ρv and the current
density J at a given point in a material are totally independent
of one another. This is no longer true in the time-varying
case. To show the connection between ρv and J, we start
by considering an arbitrary volume υ bounded by a closed
surface S (Fig. 6-14). The net positive charge contained in υ
is Q. Since, according to the law of conservation of electric
charge (Section 1-3.2), charge can neither be created nor
destroyed, the only way Q can increase is as a result of a net
inward flow of positive charge into the volume υ . By the same
token, for Q to decrease, there has to be a net outward flow
of charge from υ . The inward and outward flow of charge
constitute currents flowing across the surface S into and out
of υ , respectively. We define I as the net current flowing
across S out of υ . Accordingly, I is equal to the negative rate
I=−
Z
ν
υ
ρv d υ ,
(6.50)
d
J · ds = −
dt
S
Z
υ
ρv d υ .
(6.51)
By applying the divergence theorem given by Eq. (3.98), we
can convert the surface integral of J into a volume integral of
its divergence ∇ · J, which then gives
Z
S
J · ds =
Z
υ
∇ · J dυ = −
d
dt
Z
υ
ρv d υ .
(6.52)
For a stationary volume υ , the time derivative operates on ρv
only. Hence, we can move it inside the integral and express it
as a partial derivative of ρv :
υ
Charge density ρv
Z
dQ
d
=−
dt
dt
where ρv is the volume charge density in υ . According to
Eq. (4.12), the current I is also defined as the outward flux
of the current density J through the surface S. Hence,
Z
J
J
of change of Q:
∇ · J dυ = −
Z
υ
∂ ρv
dυ .
∂t
(6.53)
In order for the volume integrals on both sides of Eq. (6.53) to
be equal for any volume υ , their integrands have to be equal
at every point within υ . Hence,
J
S encloses ν
J
υ is
equal to the flux of the current density J through the surface S,
which in turn is equal to the rate of decrease of the charge
enclosed in υ .
Figure 6-14 The total current flowing out of a volume
∇·J = −
∂ ρv
,
∂t
(6.54)
which is known as the charge–current continuity relation, or
simply the charge continuity equation.
If the volume charge density within an elemental volume ∆υ (such as a small cylinder) is not a function of time
6-10 FREE-CHARGE DISSIPATION IN A CONDUCTOR
I1
I2
I3
Figure 6-15 Kirchhoff’s current law states that the algebraic
sum of all the currents flowing out of a junction is zero.
(i.e., ∂ ρv /∂ t = 0), it means that the net current flowing out
of ∆υ is zero or, equivalently, that the current flowing into ∆υ
is equal to the current flowing out of it. In this case, Eq. (6.54)
implies
∇ · J = 0,
(6.55)
and its integral-form equivalent [from Eq. (6.51)] is
Z
S
J · ds = 0. (Kirchhoff’s current law)
(6.56)
Let us examine the meaning of Eq. (6.56) by considering a
junction (or node) connecting two or more branches in an electric circuit. No matter how small, the junction has a volume υ
enclosed by a surface S. The junction shown in Fig. 6-15
has been drawn as a cube, and its dimensions have been
artificially enlarged to facilitate the present discussion. The
junction has six faces (surfaces), which collectively constitute
the surface S associated with the closed-surface integration
given by Eq. (6.56). For each face, the integration represents
the current flowing out through that face. Thus, Eq. (6.56) can
be cast as
∑ Ii = 0,
(Kirchhoff’s current law)
(6.57)
i
where Ii is the current flowing outward through the ith face.
For the junction of Fig. 6-15, Eq. (6.57) translates into
(I1 + I2 + I3 ) = 0. In its general form, Eq. (6.57) is an expression of Kirchhoff’s current law, which states that in an electric
circuit the sum of all the currents flowing out of a junction is
zero.
289
fluence of an externally applied electric field. These electrons,
however, are not excess charges; their charge is balanced by
an equal amount of positive charge in the atoms’ nuclei. In
other words, the conductor material is electrically neutral, and
the net charge density in the conductor is zero (ρv = 0). What
happens then if an excess free charge q is introduced at some
interior point in a conductor? The excess charge gives rise to
an electric field, which forces the charges of the host material
nearest to the excess charge to rearrange their locations, which
in turn cause other charges to move, and so on. The process
continues until neutrality is reestablished in the conductor
material and a charge equal to q resides on the conductor’s
surface.
How fast does the excess charge dissipate? To answer this
question, let us introduce a volume charge density ρvo at the
interior of a conductor and then find out the rate at which it
decays down to zero. From Eq. (6.54), the continuity equation
is given by
∂ ρv
∇·J = −
.
(6.58)
∂t
In a conductor, the point form of Ohm’s law given by
Eq. (4.73) states that J = σ E. Hence,
∂ ρv
.
(6.59)
∂t
Next, we use Eq. (6.1), ∇ · E = ρv /ε , to obtain the partial
differential equation
σ∇·E = −
∂ ρv σ
+ ρv = 0.
(6.60)
∂t
ε
Given that ρv = ρvo at t = 0, the solution of Eq. (6.60) is
ρv (t) = ρvo e−(σ /ε )t = ρvo e−t/τr
(C/m3 ),
(6.61)
where τr = ε /σ is called the relaxation time constant. We
see from Eq. (6.61) that the initial excess charge ρvo decays
exponentially at a rate τr . At t = τr , the initial charge ρvo
will have decayed to 1/e ≈ 37% of its initial value, and at
t = 3τr , it will have decayed to e−3 ≈ 5% of its initial value
at t = 0. For copper, with ε ≈ ε0 = 8.854 × 10−12 F/m and
σ = 5.8 × 107 S/m, τr = 1.53 × 10−19 s. Thus, the charge dissipation process in a conductor is extremely fast. In contrast,
the decay rate is very slow in a good insulator. For a material
like mica with ε = 6ε0 and σ = 10−15 S/m, τr = 5.31 × 104 s,
or approximately 14.8 hours.
Explain how the charge continuity equation leads to Kirchhoff’s current law.
Concept Question 6-11:
6-10 Free-Charge Dissipation in a
Conductor
We stated earlier that current flow in a conductor is realized
by the movement of loosely attached electrons under the in-
How long is the relaxation
time constant for charge dissipation in a perfect conductor? In a perfect dielectric?
Concept Question 6-12:
290
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Technology Brief 12: EMF Sensors
An electromotive force (emf) sensor is a device that can
generate an induced voltage in response to an external
stimulus. Three types of emf sensors are profiled in this
technical brief: the piezoelectric transducer, the Faraday magnetic flux sensor, and the thermocouple.
Piezoelectric Transducers
◮ Piezoelectricity is the property exhibited by certain crystals, such as quartz, that become electrically
polarized when the crystal is subjected to mechanical pressure, thereby inducing a voltage across it. ◭
The crystal consists of polar domains represented by
equivalent dipoles (Fig. TF12-1). Under the absence
of an external force, the polar domains are randomly
oriented throughout the material, but when compressive
or tensile (stretching) stress is applied to the crystal,
the polar domains align themselves along one of the
principal axes of the crystal, leading to a net polarization
(electric charge) at the crystal surfaces. Compression
and stretching generate voltages of opposite polarity.
The piezoelectric effect (piezein means to press or
squeeze in Greek) was discovered by the Curie brothers, Pierre and Paul-Jacques, in 1880, and a year later,
Lippmann predicted the converse property: If subjected
to an electric field, the crystal would change in shape.
Vemf = 0
Piezoelectric crystals are used in microphones to convert mechanical vibrations (of the crystal surface) caused
by acoustic waves into a corresponding electrical signal,
and the converse process is used in loudspeakers to
convert electrical signals into sound. In addition to having
stiffness values comparable to that of steel, some piezoelectric materials exhibit very high sensitivity to the force
applied upon them, with excellent linearity over a wide
dynamic range. They can be used to measure surface
deformations as small as nanometers (10−9 m), making them particularly attractive as positioning sensors
in scanning tunneling microscopes. As accelerometers, they can measure acceleration levels as low as
10−4 g to as high as 100 g (where g is the acceleration
due to gravity). Piezoelectric crystals and ceramics are
used in cigarette lighters and gas grills as spark generators, in clocks and electronic circuitry as precision
oscillators, in medical ultrasound diagnostic equipment
as transducers (Fig. TF12-2), and in numerous other
applications.
Faraday Magnetic Flux Sensor
According to Faraday’s law [Eq. (6.6)], the emf voltage
induced across the terminals of a conducting loop is
directly proportional to the time rate of change of the
Vemf > 0
+
_
+
_
Dipole
_
+
F=0
◮ The piezoelectric effect is a reversible (bidirectional) electromechanical process; application
of force induces a voltage across the crystal, and
conversely, application of a voltage changes the
shape of the crystal. ◭
+
_
+
_
+
_
+
_
_
+
_
+
_
+
Vemf < 0
F
F
(a) No force
_
+
_
+
_
+
(b) Compressed crystal
(c) Stretched crystal
Figure TF12-1 Response of a piezoelectric crystal to an applied force.
6-10 FREE-CHARGE DISSIPATION IN A CONDUCTOR
Case
Epoxy
potting
291
Measurement
junction
Backing
material
Cold reference junction
Copper
I
+
Coaxial cable connector
Signal wire
Ground wire
Electrodes
Piezoelectric
element
R
Vs
_
Bismuth
Wear plate
T2
T1
Figure TF12-4 Principle of the thermocouple.
Figure TF12-2 The ultrasonic transducer uses piezoelectric
crystals.
Magnet
Conducting loop
+
l
Vemf
N
_
x
u
S
Figure TF12-3 In a Faraday accelerometer, the induced emf
is directly proportional to the velocity of the loop (into and out
of the magnet’s cavity).
magnetic flux passing through the loop. For the configuration in Fig. TF12-3,
Vemf = −uB0 l,
where u = dx/dt is the velocity of the loop (into or out
of the magnet’s cavity) with the direction of u defined as
positive when the loop is moving inward into the cavity,
B0 is the magnetic field of the magnet, and l is the loop
width. With B0 and l being constant, the variation of
Vemf (t) with time t becomes a direct indicator of the time
variation of u(t). The time derivative of u(t) provides the
acceleration a(t).
Thermocouple
In 1821, Thomas Seebeck discovered that when a
junction made of two different conducting materials, such
as bismuth and copper, is heated it generates a thermally
induced emf, which we now call the Seebeck potential
VS (Fig. TF12-4). When connected to a resistor, a current
given by I = VS /R flows through the resistor.
This feature was advanced by A. C. Becquerel in
1826 as a means to measure the unknown temperature T2 of a junction relative to a temperature T1 of a
(cold) reference junction. Today, such a generator of
thermoelectricity is called a thermocouple. Initially, an
ice bath was used to maintain T1 at 0◦ C, but in today’s
temperature sensor designs, an artificial cold junction is
used instead. The artificial junction is an electric circuit
that generates a potential equal to that expected from a
reference junction at temperature T1 .
292
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Exercise 6-6: Determine (a) the relaxation time constant
and (b) the time it takes for a charge density to decay
to 1% of its initial value in quartz given that εr = 5 and
σ = 10−17 S/m.
Answer: (a) τr = 51.2 days, (b) 236 days. (See
EM
.)
6-11 Electromagnetic Potentials
Our discussion of Faraday’s and Ampère’s laws revealed two
aspects of the link between time-varying electric and magnetic
fields. We now examine the implications of this interconnection on the electric scalar potential V and the vector magnetic
potential A.
In the static case, Faraday’s law reduces to
× E = 0,
∇×
(static case)
(electrostatics)
(6.63)
In the dynamic case, Faraday’s law is
×E =−
∇×
∂B
.
∂t
(6.64)
× A, Eq. (6.64) can be expressed
In view of the relation B = ∇×
as
∂
× E = − (∇×
× A),
∇×
(6.65)
∂t
which can be rewritten as
∂A
× E+
= 0.
∇×
∂t
E = −∇V −
∂A
.
∂t
(dynamic case)
(6.70)
Equation (6.70) reduces to Eq. (6.63) in the static case.
When the scalar potential V and the vector potential A are
known, E can be obtained from Eq. (6.70), and B can be
obtained from
× A.
B = ∇×
(6.71)
Next we examine the relations between the potentials, V and A,
and their sources: the charge and current distributions ρv and J
in the time-varying case.
(6.62)
which states that the electrostatic field E is conservative.
According to the rules of vector calculus, if a vector field E
is conservative, it can be expressed as the gradient of a scalar.
Hence, in Chapter 4, we defined E as
E = −∇V .
Upon substituting Eq. (6.67) for E′ in Eq. (6.69) and then
solving for E, we have
6-11.1 Retarded Potentials
Consider the situation depicted in Fig. 6-16. A charge distribution ρv exists over a volume υ ′ embedded in a perfect
dielectric with permittivity ε . Were this a static charge distribution, then from Eq. (4.58a) the electric potential V (R) at an
observation point in space specified by the position vector R
would be
Z
1
ρv (Ri )
dυ ′,
(6.72)
V (R) =
4πε υ ′ R′
where Ri denotes the position vector of an elemental volume
∆υ ′ containing charge density ρv (Ri ), and R′ = |R − Ri | is the
distance between ∆υ ′ and the observation point. If the charge
z
(dynamic case)
(6.66)
Charge
distribution ρv
∆υ'
R'
V(R)
Let us for the moment define
E′ = E +
∂A
.
∂t
Ri
(6.67)
y
Using this definition, Eq. (6.66) becomes
′
× E = 0.
∇×
R
ν'
(6.68)
Following the same logic that led to Eq. (6.63) from Eq. (6.62),
we define
E′ = −∇V .
(6.69)
x
Figure 6-16 Electric potential V (R) due to a charge distribution ρv over a volume υ ′ .
6-11 ELECTROMAGNETIC POTENTIALS
293
distribution is time-varying, we may be tempted to rewrite
Eq. (6.72) for the dynamic case as
V (R,t) =
1
4πε
Z
υ′
ρv (Ri ,t)
dυ ′,
R′
(6.73)
but such a form does not account for “reaction time.” If V1 is
the potential due to a certain distribution ρv1 and ρv1 were to
suddenly change to ρv2 , it will take a finite amount of time
before V1 a distance R′ away changes to V2 . In other words,
V (R,t) cannot change instantaneously. The delay time is equal
to t ′ = R′ /up , where up is the velocity of propagation in the
medium between the charge distribution and the observation
point. Thus, V (R,t) at time t corresponds to ρv at an earlier
time, that is, (t − t ′). Hence, Eq. (6.73) should be rewritten as
V (R,t) =
1
4πε
Z
υ′
ρv (Ri , t − R′ /up )
dυ ′
R′
(V),
(6.74)
and V (R,t) is appropriately called the retarded scalar potential. If the propagation medium is vacuum, up is equal to the
velocity of light c.
Similarly, the retarded vector potential A(R,t) is related to
the distribution of current density J by
A(R,t) =
µ
4π
Z
υ′
J(Ri , t − R′ /up )
d υ ′ (Wb/m).
R′
(6.75)
it can be represented by a Fourier integral. In either case, if
the response of the linear system is known for all steady-state
sinusoidal excitations, the principle of superposition can be
used to determine its response to an excitation with arbitrary
time dependence. Thus, the sinusoidal response of the system
constitutes a fundamental building block that can be used
to determine the response due to a source described by an
arbitrary function of time. The term time-harmonic is often
used in this context as a synonym for “steady-state sinusoidal
time-dependent.”
In this subsection, we derive expressions for the scalar and
vector potentials due to time-harmonic sources. Suppose that
ρv (Ri ,t) is a sinusoidal-time function with angular frequency
ω given by
ρv (Ri ,t) = ρv (Ri ) cos(ω t + φ ).
Phasor analysis, which was first introduced in Section 1-7
and then used extensively in Chapter 2 to study wave propagation on transmission lines, is a useful tool for analyzing
time-harmonic scenarios. A time harmonic charge distribution
ρv (Ri ,t) is related to its phasor ρ̃v (Ri ) as
ρv (Ri ,t) = Re ρ̃v (Ri ) e jω t ,
(6.77)
Comparison of Eqs. (6.76) and (6.77) shows that in the present
case ρ̃v (Ri ) = ρv (Ri ) e jφ .
Next, we express the retarded charge density
ρv (Ri ,t − R′ /up ) in phasor form by replacing t with (t − R′ /up )
in Eq. (6.77):
h
i
′
ρv (Ri ,t − R′ /up ) = Re ρ̃v (Ri ) e jω (t−R /up )
i
h
′
= Re ρ̃v (Ri ) e− jω R /up e jω t
i
h
′
(6.78)
= Re ρ̃v (Ri ) e− jkR e jω t ,
This expression is obtained by extending the expression for the
magnetostatic vector potential A(R) given by Eq. (5.65) to the
time-varying case.
6-11.2 Time-Harmonic Potentials
The expressions given by Eqs. (6.74) and (6.75) for the
retarded scalar and vector potentials are valid under both static
and dynamic conditions and for any type of time dependence
of the source functions ρv and J. Because V and A depend
linearly on ρv and J, and as E and B depend linearly on V
and A, the relationships interconnecting all of these quantities
obey the rules of linear systems. When analyzing linear
systems, we can take advantage of sinusoidal-time functions
to determine the system’s response to a source with arbitrary
time dependence. As was noted in Section 1-7, if the time
dependence is described by a (nonsinusoidal) periodic time
function, it can always be expanded into a Fourier series of
sinusoidal components, and if the time function is nonperiodic,
(6.76)
where
k=
ω
up
(6.79)
is called the wavenumber or phase constant of the propagation
medium. (In general, the phase constant is denoted by the
symbol “β ”, but for lossless dielectric media, it is commonly
denoted by the symbol “k” and called the wavenumber.)
e (R) of the time function
Similarly, we define the phasor V
V (R,t) according to
i
h
e (R) e jω t .
V (R,t) = Re V
(6.80)
294
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Using Eqs. (6.78) and (6.80) in Eq. (6.74) gives
#
′
ρ̃v (Ri ) e− jkR jω t
′
e dυ .
R′
υ′
(6.81)
By equating the quantities inside the square brackets on both
sides of Eq. (6.81) and canceling the common e jω t factor, we
obtain the phasor-domain expression
h
i
Re Ve (R) e jω t = Re
1
Ve (R) =
4πε
Z
"
1
4πε
Z
′
ρ̃v (Ri ) e− jkR
dυ ′
R′
υ′
(V).
h
i
e
A(R,t) = Re A(R)
e jω t
µ
e
A(R)
=
4π
Z
υ′
′
e
J(Ri ) e− jkR
dυ ′,
R′
e
e = 1 ∇×
× A.
H
µ
(6.87)
In a nonconducting medium with ε = 16ε0 and µ = µ0 , the
electric field intensity of an electromagnetic wave is
(6.83)
(6.84)
(6.85)
Recall that differentiation in the time domain is equivalent
to multiplication by jω in the phasor domain, and in a
nonconducting medium (J = 0), so Ampère’s law given by
Eq. (6.41) becomes
or
e = − 1 ∇×
e
× E.
H
jω µ
(6.82)
where e
J(Ri ) is the phasor function corresponding to J(Ri ,t).
e is given by
e corresponding to A
The magnetic field phasor H
e = jωε E
e
×H
∇×
or
e
e = − jω µ H
×E
∇×
Example 6-8: Relating E to H
For any given charge distribution, Eq. (6.82) can be used to
e (R). Then the resultant expression can be used in
compute V
Eq. (6.80) to find V (R,t). Similarly, the expression for A(R,t)
given by Eq. (6.75) can be transformed into
with
e and H.
e The phasor vectors E
e and H
e also
to determine both E
are related by the phasor form of Faraday’s law:
e = 1 ∇×
e
× H.
E
jωε
(6.86)
Hence, given a time-harmonic, current-density distribution
with phasor e
J, Eqs. (6.84) to (6.86) can be used successively
E(z,t) = x̂ 10 sin(1010t − kz)
(V/m).
(6.88)
Determine the associated magnetic field intensity H and find
the value of k.
e
Solution: We begin by finding the phasor E(z)
of E(z,t).
Since E(z,t) is given as a sine function and phasors are defined
in this book with reference to the cosine function, we rewrite
Eq. (6.88) as
E(z,t) = x̂ 10 cos(1010t − kz − π /2)
h
i
e e jω t ,
= Re E(z)
(V/m)
(6.89)
with ω = 1010 (rad/s) and
e = x̂ 10e− jkze− jπ /2
E(z)
= −x̂ j10e− jkz .
(6.90)
e
To find both H(z)
and k, we will perform a “circle”: We will
e in Faraday’s law to find H(z);
e
use the given expression for E(z)
e
e
then we will use H(z)
in Ampère’s law to find E(z),
which
e
we will then compare with the original expression for E(z);
and the comparison will yield the value of k. Application of
Eq. (6.87) gives
e
e = − 1 ∇×
×E
H(z)
jω µ
ŷ
ẑ
∂ /∂ y ∂ /∂ z
0
0
∂
1
ŷ (− j10e− jkz )
=−
jω µ
∂z
=−
1
jω µ
= −ŷ j
x̂
∂ /∂ x
− j10e− jkz
10k − jkz
e
.
ωµ
(6.91)
6-11 ELECTROMAGNETIC POTENTIALS
295
e
e
So far, we have used Eq. (6.90) for E(z)
to find H(z),
but k
e
remains unknown. To find k, we use H(z) in Eq. (6.86) to find
e
E(z):
e
e = 1 ∇×
×H
E(z)
jωε
10k − jkz
∂
1
−j
e
=
−x̂
jωε
∂z
ωµ
= −x̂ j
10k2
ω 2 µε
e− jkz .
= ŷ 0.11 sin(1010t − 133z)
(6.92)
Equating Eqs. (6.90) and (6.92) leads to
2
With k known, the instantaneous magnetic field intensity is
then given by
i
h
e e jω t
H(z,t) = Re H(z)
10k − jkz jω t
= Re −ŷ j
e
e
ωµ
(A/m).
(6.94)
We note that k has the same expression as the phase constant
of a lossless transmission line [Eq. (2.49)].
Exercise 6-7: The magnetic field intensity of an elec-
tromagnetic wave propagating in a lossless medium with
ε = 9ε0 and µ = µ0 is
2
k = ω µε ,
or
H(z,t) = x̂ 0.3 cos(108t − kz + π /4)
√
k = ω µε
√
= 4ω µ0 ε0
(A/m).
Find E(z,t) and k.
4ω
4 × 1010
= 133 (rad/m).
=
=
c
3 × 108
Answer: E(z,t) = −ŷ 37.7 cos(108t − z + π /4) (V/m);
(6.93)
k = 1 (rad/m). (See
EM
.)
Chapter 6 Summary
Concepts
• Faraday’s law states that a voltage is induced across
the terminals of a loop if the magnetic flux linking its
surface changes with time.
• In an ideal transformer, the ratios of the primary to
secondary voltages, currents, and impedances are governed by the turns ratio.
• Displacement current accounts for the “apparent” flow
of charges through a dielectric. In reality, charges of
opposite polarity accumulate along the two ends of
a dielectric, giving the appearance of current flow
through it.
• Boundary conditions for the electromagnetic fields at
the interface between two different media are the same
for both static and dynamic conditions.
• The charge continuity equation is a mathematical statement of the law of conservation of electric charge.
• Excess charges in the interior of a good conductor dissipate very quickly; through a rearrangement process,
the excess charge is transferred to the surface of the
conductor.
• In the dynamic case, the electric field E is related to
both the scalar electric potential V and the magnetic
vector potential A.
• The retarded scalar and vector potentials at a given
observation point take into account the finite time
required for propagation between their sources, the
charge and current distributions, and the location of the
observation point.
296
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
Mathematical and Physical Models
Faraday’s Law
Vemf = −
EM Potentials
dΦ
d
=−
dt
dt
Transformer
tr
Vemf
= −N
Motional
m
Vemf
=
Z
C
Z
S
∂A
E = −∇V −
∂t
×A
B = ∇×
tr
m
B · ds = Vemf
+ Vemf
Z
∂B
· ds
S ∂t
Current Density
(N loops)
× B) · dl
(u×
Jc = σ E
Displacement
Jd =
∂D
∂t
Conductor Charge Dissipation
ρv (t) = ρvo e−(σ /ε )t = ρvo e−t/τr
Charge–Current Continuity
∇·J = −
Conduction
∂ ρv
∂t
Important Terms
Provide definitions or explain the meaning of the following terms:
charge continuity equation
charge dissipation
displacement current Id
electromagnetic induction
electromotive force Vemf
retarded potential
tr
transformer emf Vemf
wavenumber k
Faraday’s law
Kirchhoff’s current law
Lenz’s law
m
motional emf Vemf
relaxation time constant
PROBLEMS
R2
Sections 6-1 to 6-6: Faraday’s Law and its Applications
I
∗ 6.1 The switch in the bottom loop of Fig. P6.1 is closed at
t = 0 and then opened at a later time t1 . What is the direction of
the current I in the top loop (clockwise or counterclockwise)
at each of these two times?
6.2 The loop in Fig. P6.2 is in the x–y plane and
B = ẑB0 sin ω t with B0 positive. What is the direction of I
(φ̂φ or −φ̂φ) at:
(a) t = 0
(b) ω t = π /4
(c) ω t = π /2
6.3 A coil consists of 100 turns of wire wrapped around
a square frame of sides 0.25 m. The coil is centered at the
origin with each of its sides parallel to the x or y axis. Find the
∗ Answer(s) available in Appendix E.
t = t1
t=0
R1
Figure P6.1 Loops of Problem 6.1.
induced emf across the open-circuited ends of the coil if the
magnetic field is given by
∗ (a) B = ẑ 20e−3t (T)
(b) B = ẑ 20 cos x cos 103t (T)
(c) B = ẑ 20 cos x sin 2y cos 103t (T)
PROBLEMS
297
z
z
10 cm
R
I(t)
y
Vemf
10 cm
5 cm
I
x
y
Figure P6.2 Loop of Problem 6.2.
x
6.4 A stationary conducting loop with an internal resistance
of 0.5 Ω is placed in a time-varying magnetic field. When the
loop is closed, a current of 5 A flows through it. What will the
current be if the loop is opened to create a small gap and a 2 Ω
resistor is connected across its open ends?
Figure P6.6 Loop coplanar with long wire (Problem 6.6).
z
∗ 6.5 A circular-loop TV antenna with 0.02 m2 area is in
2c
m
the presence of a uniform-amplitude 300 Mhz signal. When
oriented for maximum response, the loop develops an emf with
a peak value of 30 (mV). What is the peak magnitude of B of
the incident wave?
B
3 cm
B
y
6.6 The square loop shown in Fig. P6.6 is coplanar with a
long, straight wire carrying a current
4
I(t) = 5 cos(2π × 10 t)
(A).
φ(t)
ω
x
Figure P6.7 Rotating loop in a magnetic field (Problem 6.7).
(a) Determine the emf induced across a small gap created in
the loop.
(b) Determine the direction and magnitude of the current that
would flow through a 4 Ω resistor connected across the
gap. The loop has an internal resistance of 1 Ω.
∗ 6.7 The rectangular conducting loop shown in Fig. P6.7
rotates at 6,000 revolutions per minute in a uniform magnetic
flux density given by
B = ŷ 50
(mT).
Determine the current induced in the loop if its internal
resistance is 0.5 Ω.
6.8 The transformer shown in Fig. P6.8 consists of a long
wire coincident with the z axis carrying a current I = I0 cos ω t,
coupling magnetic energy to a toroidal coil situated in the x–y
plane and centered at the origin. The toroidal core uses iron
material with relative permeability µr , around which 100 turns
of a tightly wound coil serves to induce a voltage Vemf , as
shown in the figure.
(a) Develop an expression for Vemf .
(b) Calculate Vemf for f = 60 Hz, µr = 4000, a = 5 cm,
b = 6 cm, c = 2 cm, and I0 = 50 A.
6.9 A rectangular conducting loop 5 cm × 10 cm with a
small air gap in one of its sides is spinning at 7200 revolutions
per minute. If the field B is normal to the loop axis and its
298
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
z
z
10 cm
R
I
u
I1 = 10 A
N
Vemf
I2
20 cm
b
c
a
y
u
R
y0
Iron core with μr
x
Figure P6.8 Problem 6.8.
Figure P6.11 Moving loop of Problem 6.11.
magnitude is 6 × 10−6 T, what is the peak voltage induced
across the air gap?
per minute in a uniform magnetic flux density B0 = 0.4 T,
determine the amplitude of the current generated in the light
bulb.
∗ 6.10 A 50-cm long metal rod rotates about the z axis at
90 revolutions per minute with end 1 fixed at the origin
as shown in Fig. P6.10. Determine the induced emf V12 if
B = ẑ 2 × 10−4 T.
6.13 The circular, conducting disk shown in Fig. P6.13 lies
in the x–y plane and rotates with uniform angular velocity ω
about the z axis. The disk is of radius a and is present in a
uniform magnetic flux density B = ẑB0 . Obtain an expression
for the emf induced at the rim relative to the center of the disk.
z
z
V
B
1
y
2
ω
ω
x
a
y
Figure P6.10 Rotating rod of Problem 6.10.
x
6.11 The loop shown in Fig. P6.11 moves away from a
wire carrying a current I1 = 10 A at a constant velocity
u = ŷ7.5 (m/s). If R = 10 Ω and the direction of I2 is as defined
in the figure, find I2 as a function of y0 , which is the distance
between the wire and the loop. Ignore the internal resistance
of the loop.
∗ 6.12 The electromagnetic generator shown in Fig. 6-12 is
connected to an electric bulb with a resistance of 150 Ω. If
the loop area is 0.1 m2 and it rotates at 3,600 revolutions
Figure P6.13
Rotating circular disk in a magnetic field
(Problem 6.13).
Section 6-7: Displacement Current
6.14 The plates of a parallel-plate capacitor have areas of
10 cm2 each and are separated by 2 cm. The capacitor is filled
PROBLEMS
299
with a dielectric material with ε = 4ε0 , and the voltage across it
is given by V (t) = 30 cos2π × 106t (V). Find the displacement
current.
∗ 6.15 A coaxial capacitor of length l = 6 cm uses an insulat-
ing dielectric material with εr = 9. The radii of the cylindrical
conductors are 0.5 cm and 1 cm. If the voltage applied across
the capacitor is
V (t) = 50 sin(120π t)
(V),
∗ 6.17 In wet soil characterized by σ = 10−2 (S/m), µ = 1,
r
and εr = 36, at what frequency is the conduction current density equal in magnitude to the displacement current density?
6.18 An electromagnetic wave propagating in seawater has
an electric field with a time variation given by E = ẑE0 cos ω t.
If the permittivity of water is 81ε0 and its conductivity is
4 (S/m), find the ratio of the magnitudes of the conduction
current density to displacement current density at each of the
following frequencies:
(a) 1 kHz
what is the displacement current?
∗ (b) 1 MHz
6.16 The parallel-plate capacitor shown in Fig. P6.16 is filled
with a lossy dielectric material of relative permittivity εr and
conductivity σ . The separation between the plates is d and
each plate is of area A. The capacitor is connected to a timevarying voltage source V (t).
(c) 1 GHz
(d) 100 GHz
Sections 6-9 and 6-10: Continuity Equation and Charge
Dissipation
6.19 At t = 0, charge density ρv0 was introduced into the
interior of a material with a relative permittivity εr = 9. If at
t = 1 µ s the charge density has dissipated down to 10−3 ρv0 ,
what is the conductivity of the material?
I
A
∗ 6.20 If the current density in a conducting medium is given
by
V(t)
ε, σ
d
J(x, y, z;t) = (x̂z − ŷ4y2 + ẑ2x) cos ω t,
determine the corresponding charge distribution ρv (x, y, z;t).
6.21 In a certain medium, the direction of current density J
points in the radial direction in cylindrical coordinates, and its
magnitude is independent of both φ and z. Determine J given
that the charge density in the medium is
ρv = ρ0 r cos ω t (C/m3 ).
Figure P6.16
Parallel-plate capacitor containing a lossy
dielectric material (Problem 6.16).
(a) Obtain an expression for Ic , the conduction current flowing between the plates inside the capacitor, in terms of the
given quantities.
(b) Obtain an expression for Id , the displacement current
flowing inside the capacitor.
(c) Based on your expressions for parts (a) and (b), give an
equivalent-circuit representation for the capacitor.
(d) Evaluate the values of the circuit elements for A = 4 cm2 ,
d = 0.5 cm, εr = 4, σ = 2.5 (S/m), and
V (t) = 10 cos(3π × 103t) (V).
6.22 If we were to characterize how good a material is as an
insulator by its resistance to dissipating charge, which of the
following two materials is the better insulator?
Dry Soil:
Fresh Water:
εr = 2.5,
εr = 80,
σ = 10−4 (S/m)
σ = 10−3 (S/m)
Sections 6-11: Electromagnetic Potentials
6.23 The electric field of an electromagnetic wave propagating in air is given by
E(z,t) = x̂4 cos(6 × 108t − 2z)
+ ŷ3 sin(6 × 108t − 2z)
Find the associated magnetic field H(z,t).
(V/m).
300
CHAPTER 6 MAXWELL’S EQUATIONS FOR TIME-VARYING FIELDS
∗ 6.24 The magnetic field in a dielectric material with ε = 4ε ,
0
µ = µ0 , and σ = 0 is given by
H(y,t) = x̂5 cos(2π × 107t + ky)
(A/m).
Find k and the associated electric field E.
6.25 Given an electric field
E = x̂E0 sin ay cos(ω t − kz),
where E0 , a, ω , and k are constants, find H.
∗ 6.26 The electric field radiated by a short dipole antenna is
given in spherical coordinates by
2 × 10−2
sin θ cos(6π × 108t − 2π R) (V/m).
E(R, θ ;t) = θ̂θ
R
Find H(R, θ ;t).
6.27 A Hertzian dipole is a short conducting wire carrying
an approximately constant current over its length l. If such a
dipole is placed along the z axis with its midpoint at the origin
and if the current flowing through it is i(t) = I0 cos ω t, find the
following:
e θ , φ ) at an observation
(a) The retarded vector potential A(R,
point Q(R, θ , φ ) in a spherical coordinate system.
e θ , φ ).
(b) The magnetic field phasor H(R,
Assume l to be sufficiently small so that the observation point
is approximately equidistant to all points on the dipole; that is,
assume R′ ≈ R.
6.28 In free space, the magnetic field is given by
φ
H = φ̂
36
cos(6 × 109t − kz)
r
(mA/m).
∗ (a) Determine k.
(b) Determine E.
(c) Determine Jd .
6.29 The magnetic field in a given dielectric medium is given
by
H = ŷ 6 cos2z sin(2 × 107t − 0.1x)
(A/m),
where x and z are in meters.
(a) Determine E.
(b) Determine the displacement current density Jd .
(c) Determine the charge density ρv .
Chapter
7
Plane-Wave Propagation
Chapter Contents
7-1
7-2
7-3
TB13
7-4
TB14
7-5
7-6
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Unbounded EM Waves, 302
Time-Harmonic Fields, 303
Plane-Wave Propagation in Lossless Media, 304
Wave Polarization, 309
RFID Systems, 313
Plane-Wave Propagation in Lossy Media, 317
Liquid Crystal Display (LCD), 320
Current Flow in a Good Conductor, 325
Electromagnetic Power Density, 328
Chapter 7 Summary, 331
Problems, 333
1. Describe mathematically the electric and magnetic fields
of TEM waves.
2. Describe the polarization properties of an EM wave.
3. Relate the propagation parameters of a wave to the
constitutive parameters of the medium.
4. Characterize the flow of current in conductors and use it
to calculate the resistance of a coaxial cable.
5. Calculate the rate of power carried by an EM wave in both
lossless and lossy media.
301
Unbounded EM Waves
Radiating
antenna
I
t was established in Chapter 6 that a time-varying electric field produces a magnetic field and, conversely, a timevarying magnetic field produces an electric field. This cyclic
pattern often results in electromagnetic (EM) waves propagating through free space and in material media. When a
wave propagates through a homogeneous medium without
interacting with obstacles or material interfaces, it is said to be
unbounded. Light waves emitted by the sun and radio transmissions by antennas are good examples. Unbounded waves
may propagate in both lossless and lossy media. Waves propagating in a lossless medium (e.g., air and perfect dielectrics)
are similar to those on a lossless transmission line in that
they do not attenuate. When propagating in a lossy medium
(material with nonzero conductivity, such as water), part of the
power carried by an EM wave gets converted into heat. A wave
produced by a localized source, such as an antenna, expands
outwardly in the form of a spherical wave, as depicted in
Fig. 7-1(a). Even though an antenna may radiate more energy
along some directions than along others, the spherical wave
travels at the same speed in all directions. To an observer
very far away from the source, however, the wavefront of the
spherical wave appears approximately planar, as if it were
part of a uniform plane wave with identical properties at all
points in the plane tangent to the wavefront (Fig. 7-1(b)).
Plane waves are easily described using a Cartesian coordinate
system, which is mathematically easier to work with than
the spherical coordinate system needed to describe spherical
waves.
When a wave propagates along a material structure, it is said
to be guided. The Earth’s surface and ionosphere constitute
parallel boundaries of a natural structure capable of guiding
short-wave radio transmissions in the HF band∗ (3 to 30 MHz);
indeed, the ionosphere is a good reflector at these frequencies, thereby allowing the waves to zigzag between the two
boundaries (Fig. 7-2). When we discussed wave propagation
on a transmission line in Chapter 2, we dealt with voltages and
currents. For a transmission-line circuit such as that shown in
Fig. 7-3, the ac voltage source excites an incident wave that
travels down the coaxial line toward the load, and unless the
load is matched to the line, part (or all) of the incident wave
is reflected back toward the generator. At any point on the
line, the instantaneous total voltage v(z,t) is the sum of the
incident and reflected waves—both of which vary sinusoidally
with time. Associated with the voltage difference between the
inner and outer conductors of the coaxial line is a radial electric
field E(z,t) that exists in the dielectric material between the
conductors, and since v(z,t) varies sinusoidally with time, so
does E(z,t). Furthermore, the current flowing through the inner
∗ See
Spherical
wavefront
(a) Spherical wave
Uniform plane wave
Aperture
Observer
(b) Plane-wave approximation
Figure 7-1 Waves radiated by an EM source, such as a light
bulb or an antenna, have spherical wavefronts, as in (a); to a
distant observer, however, the wavefront across the observer’s
aperture appears approximately planar, as in (b).
Ionosphere
Transmitter
Earth's surface
Figure 7-2 The atmospheric layer bounded by the ionosphere
at the top and the Earth’s surface at the bottom forms a guiding
structure for the propagation of radio waves in the HF band.
Fig. 1-17.
302
7-1 TIME-HARMONIC FIELDS
303
following form in the phasor domain.
H
H
H
H
Rg
Vg
E
E
e = ρ̃v /ε ,
∇·E
RL
coaxial transmission line consists of time-varying electric and
magnetic fields in the dielectric medium between the inner and
outer conductors.
conductor induces an azimuthal magnetic field H(z,t) in the
dielectric material surrounding it. These coupled fields, E(z,t)
and H(z,t), constitute an electromagnetic wave. Thus, we can
model wave propagation on a transmission line either in terms
of the voltages across the line and the currents in its conductors
or in terms of the electric and magnetic fields in the dielectric
medium between the conductors.
In this chapter, we focus our attention on wave propagation
in unbounded media. Unbounded waves have many practical
applications in science and engineering. We consider both
lossless and lossy media. Even though strictly speaking uniform plane waves cannot exist, we study them in this chapter
to develop a physical understanding of wave propagation in
lossless and lossy media. In Chapter 8, we examine how both
planar and spherical waves are reflected by and transmitted
through boundaries between dissimilar media. The proc esses
of radiation and reception of waves by antennas are treated in
Chapter 9.
7-1 Time-Harmonic Fields
Time-varying electric and magnetic fields (E, D, B, and H) and
their sources (the charge density ρv and current density J) generally depend on the spatial coordinates (x, y, z) and the time
variable t. However, if their time variation is sinusoidal with
angular frequency ω , then these quantities can be represented
by a phasor that depends on (x, y, z) only. The vector phasor
e y, z) and the instantaneous field E(x, y, z;t) it describes are
E(x,
related as
i
h
e y, z) e jω t .
(7.1)
E(x, y, z;t) = Re E(x,
Similar definitions apply to D, B, and H, as well as to ρv
and J. For a linear, isotropic, and homogeneous medium
with electrical permittivity ε , magnetic permeability µ , and
conductivity σ , Maxwell’s equations (6.1) to (6.4) assume the
e = − jω µ H,
e
×E
∇×
(7.2b)
e
e =e
×H
∇×
J + jωε E.
(7.2d)
e = 0,
∇·H
Figure 7-3 A guided electromagnetic wave traveling in a
(7.2a)
(7.2c)
To derive these equations, we used D = ε E and B = µ H as well
as the fact that, for time-harmonic quantities, differentiation in
the time domain corresponds to multiplication by jω in the
phasor domain. These equations are the starting point for the
subject matter treated in this chapter.
7-1.1 Complex Permittivity
In a medium with conductivity σ , the conduction current
e by e
e Assuming no other current
density e
J is related to E
J = σ E.
flows in the medium, Eq. (7.2d) may be written as
e =e
e (7.3)
e = (σ + jωε )E
e = jω ε − j σ E.
×H
∇×
J + jωε E
ω
By defining the complex permittivity εc as
εc = ε − j
σ
,
ω
(7.4)
Eq. (7.3) can be rewritten as
e = jωεc E.
e
×H
∇×
(7.5)
In a source-free medium, ρv = 0. Hence, Maxwell’s equations
become
e = 0,
∇·E
(7.6a)
e
e = − jω µ H,
×E
∇×
(7.6b)
e
e = jωεc E.
×H
∇×
(7.6d)
e = 0,
∇·H
(7.6c)
The complex permittivity εc given by Eq. (7.4) is often written
in terms of a real part ε ′ and an imaginary part ε ′′ . Thus,
εc = ε − j
σ
= ε ′ − jε ′′ ,
ω
(7.7)
with
ε′ = ε,
σ
ε ′′ = .
ω
(7.8a)
(7.8b)
304
CHAPTER 7 PLANE-WAVE PROPAGATION
For a lossless medium with σ = 0, it follows that ε ′′ = 0 and
εc = ε ′ = ε .
7-2 Plane-Wave Propagation in Lossless
Media
7-1.2 Wave Equations
The properties of an electromagnetic wave, such as its phase
velocity up and wavelength λ , depend on the angular frequency ω and the medium’s three constitutive parameters: ε ,
µ , and σ . If the medium is nonconducting (σ = 0), the wave
does not suffer any attenuation as it travels; hence, the medium
is said to be lossless. Because in a lossless medium εc = ε ,
Eq. (7.14) becomes
e and H
e and then solve
Next, we derive wave equations for E
e
e as a function
them to obtain explicit expressions for E and H
of the spatial variables (x, y, z). To that end, we start by taking
the curl of both sides of Eq. (7.6b) to get
e
e = − jω µ (∇×
× H).
× (∇×
× E)
∇×
(7.9)
Upon substituting Eq. (7.6d) into Eq. (7.9), we obtain
e = − jω µ ( jωεc E)
e = ω 2 µεc E.
e
× (∇×
× E)
∇×
(7.10)
e is
From Eq. (3.113), we know that the curl of the curl of E
e = ∇(∇ · E)
e − ∇2E,
e
× (∇×
× E)
∇×
(7.11)
e is the Laplacian of E,
e which in Cartesian coordiwhere ∇2 E
nates is given by
2
∂
∂2
∂2 e
2e
E.
(7.12)
+
+
∇ E=
∂ x2 ∂ y2 ∂ z2
In view of Eq. (7.6a), the use of Eq. (7.11) in Eq. (7.10) gives
e = 0,
∇ E + ω µεc E
(7.13)
γ 2 = −ω 2 µεc ,
(7.14)
2e
2
e By
which is known as the homogeneous wave equation for E.
defining the propagation constant γ as
Eq. (7.13) can be written as
e − γ 2E
e
e = 0. (wave equation for E)
∇2 E
(7.15)
To derive Eq. (7.15), we took the curl of both sides of
e and
Eq. (7.6b) and then we used Eq. (7.6d) to eliminate H
e
obtain an equation in E only. If we reverse the process, that
is, if we start by taking the curl of both sides of Eq. (7.6d) and
e we obtain a wave equation
then use Eq. (7.6b) to eliminate E,
e
for H:
e − γ 2H
e = 0.
∇2 H
e
(wave equation for H)
(7.16)
e and H
e are of the same form, so
Since the wave equations for E
are their solutions.
γ 2 = −ω 2 µε .
(7.17)
For lossless media, it is customary to define the wavenumber
k as
√
k = ω µε .
(7.18)
In view of Eq. (7.17), γ 2 = −k2 and Eq. (7.15) becomes
7-2.1
e + k2 E
e = 0.
∇2 E
(7.19)
Uniform Plane Waves
For an electric field phasor defined in Cartesian coordinates as
e = x̂Eex + ŷEey + ẑEez ,
E
substitution of Eq. (7.12) into Eq. (7.19) gives
2
∂
∂2
∂2
(x̂Eex + ŷEey + ẑEez )
+
+
∂ x2 ∂ y2 ∂ z2
+ k2 (x̂Eex + ŷEey + ẑEez ) = 0.
(7.20)
(7.21)
To satisfy Eq. (7.21), each vector component on the left-hand
side of the equation must vanish. Hence,
2
∂
∂2
∂2
2 e
+
+
+
k
Ex = 0,
(7.22)
∂ x2 ∂ y2 ∂ z2
and similar expressions apply to Eey and Eez .
◮ A uniform plane wave is characterized by electric and
magnetic fields that have uniform properties at all points
across an infinite plane. ◭
If this happens to be the x–y plane, then E and H do not vary
with x or y. Hence, ∂ Eex /∂ x = 0 and ∂ Eex /∂ y = 0, so Eq. (7.22)
reduces to
d 2 Eex
+ k2 Eex = 0.
(7.23)
dz2
7-2
PLANE-WAVE PROPAGATION IN LOSSLESS MEDIA
ex , and H
ey . The remaining
Similar expressions apply to Eey , H
e and H
e are zero; that is, Eez = H
ez = 0. To show
components of E
that Eez = 0, let us consider the z component of Eq. (7.6d),
!
ey ∂ H
ex
∂H
= ẑ jωε Eez .
ẑ
−
(7.24)
∂x
∂y
ey /∂ x = ∂ H
ex /∂ y = 0, it follows that Eez = 0. A similar
Since ∂ H
ez = 0.
examination involving Eq. (7.6b) reveals that H
◮ This means that a plane wave has no electric-field or
magnetic-field components along its direction of propagation. ◭
For the phasor quantity Eex , the general solution of the
ordinary differential equation given by Eq. (7.23) is
+ − jkz
− jkz
Eex (z) = Eex+ (z) + Eex− (z) = Ex0
e
+ Ex0
e ,
+
where Ex0
and
are constants to be determined from boundary conditions. The solution given by Eq. (7.25) is similar
e (z) given by
in form to the solution for the phasor voltage V
Eq. (2.54a) for the lossless transmission line. The first term
in Eq. (7.25), containing the negative exponential e− jkz , repre+
sents a wave with amplitude Ex0
traveling in the +z direction.
Likewise, the second term (with e jkz ) represents a wave with
−
amplitude Ex0
traveling in the −z direction. Assume for the
e only has a component along x (i.e., Eey = 0)
time being that E
e
and that Ex is associated with a wave traveling in the +z
−
direction only (i.e., Ex0
= 0). Under these conditions,
e = x̂Eex+ (z) = x̂E + e− jkz .
E(z)
x0
(7.26)
e associated with this wave, we
To find the magnetic field H
e
e
apply Eq. (7.6b) with Ey = Ez = 0:
e=
×E
∇×
ŷ
∂
∂y
0
ẑ
∂
∂z
0
ex + ŷH
ey + ẑH
ez ).
= − jω µ (x̂H
(7.27)
For a uniform plane wave traveling in the +z direction,
∂ Ex+ (z)/∂ x
= ∂ Ex+ (z)/∂ y
ez =
H
∂ Eex+ (z)
1
,
− jω µ
∂z
= 0.
(7.28a)
(7.28b)
∂ Ex+ (z)
1
= 0.
− jω µ
∂y
(7.28c)
Use of Eq. (7.26) in Eq. (7.28b) gives
ey (z) = k E + e− jkz = H + e− jkz ,
H
y0
ω µ x0
+
ey (z) and is given by
where Hy0
is the amplitude of H
+
Hy0
=
k +
E .
ω µ x0
(7.29)
(7.30)
For a wave traveling from the source toward the load on a
transmission line, the amplitudes of its voltage and current
phasors, V0+ and I0+ , are related by the characteristic impedance of the line, Z0 . A similar connection exists between the
electric and magnetic fields of an electromagnetic wave. The
intrinsic impedance of a lossless medium is defined as
ωµ
ωµ
η=
=
= √
k
ω µε
r
µ
ε
(Ω),
(7.31)
where we used the expression for k given by Eq. (7.18).
In view of Eq. (7.31), the electric and magnetic fields of a
+z-propagating plane wave with E field along x̂ are
e = x̂Eex+ (z) = x̂E + e− jkz ,
E(z)
x0
+
Ex0
− jkz
e+
e = ŷ Ex (z) = ŷ
H(z)
e
η
η
(7.32a)
.
(7.32b)
◮ The electric and magnetic fields of a plane wave are
perpendicular to each other, and both are perpendicular
to the direction of wave travel (Fig. 7-4). These attributes
qualify the wave as transverse electromagnetic (TEM). ◭
Other examples of TEM waves include waves traveling on
coaxial transmission lines (E is along r̂, H is along φ̂φ, and
the direction of travel is along ẑ) and spherical waves radiated
by antennas.
+
In the general case, Ex0
is a complex quantity with magni+
tude |Ex0 | and phase angle φ + . That is,
+
+
+ jφ
Ex0
= |Ex0
|e .
Hence, Eq. (7.27) gives
ex = 0,
H
ey =
H
(7.25)
−
Ex0
x̂
∂
∂x
Eex+ (z)
305
(7.33)
The instantaneous electric and magnetic fields therefore are
h
i
e e jω t = x̂|E + | cos(ω t − kz + φ +) (V/m),
E(z,t) = Re E(z)
x0
(7.34a)
306
CHAPTER 7 PLANE-WAVE PROPAGATION
Example 7-1: EM Plane Wave in Air
x
This example is analogous to the “Sound Wave in Water”
problem given by Example 1-1.
The electric field of a 1 MHz plane wave traveling in the
+z direction in air points along the x direction. If this field
reaches a peak value of 1.2π (mV/m) at t = 0 and z = 50 m,
obtain expressions for E(z,t) and H(z,t). Then plot them as a
function of z at t = 0.
E
kˆ
z
H
y
Figure 7-4 A transverse electromagnetic (TEM) wave propa-
Solution: At f = 1 MHz, the wavelength in air is
gating in the direction k̂ = ẑ. For all TEM waves, k̂ is parallel to
× H.
E×
λ=
3 × 108
c
= 300 m,
=
f
1 × 106
and the corresponding wavenumber is k = (2π /300) (rad/m).
The general expression for an x-directed electric field traveling
in the +z direction is given by Eq. (7.34a) as
and
+
i
h
e e jω t = ŷ |Ex0 | cos(ω t − kz+ φ + )
H(z,t) = Re H(z)
η
(A/m).
(7.34b)
Because E(z,t) and H(z,t) exhibit the same functional dependence on z and t, they are said to be in phase; when the
amplitude of one of them reaches a maximum, the amplitude
e and H
e are in phase is
of the other does so too. The fact that E
characteristic of waves propagating in lossless media.
From the material on wave motion presented in Section 1-4,
we deduce that the phase velocity of the wave is
up =
1
ω
ω
=√
= √
k
ω µε
µε
(7.35)
−
2π × 50
+ φ+ = 0
300
or
φ+ =
π
.
3
2π z π
+
E(z,t) = x̂ 1.2π cos 2π × 106t −
300 3
(mV/m),
and from Eq. (7.34b), we have
E(z,t)
η0
2π z π
6
= ŷ 10 cos 2π × 10 t −
+
300 3
H(z,t) = ŷ
(m).
(7.36)
In a vacuum, ε = ε0 and µ = µ0 , and the phase velocity up and
the intrinsic impedance η given by Eq. (7.31) are
1
= 3 × 108
(m/s),
µ0 ε0
r
µ0
η = η0 =
= 377 (Ω) ≈ 120π (Ω),
ε0
up = c = √
The field E(z,t) is maximum when the argument of the cosine
function equals zero or a multiple of 2π . At t = 0 and z = 50 m,
this condition yields
Hence,
(m/s),
and its wavelength is
up
2π
λ=
=
k
f
+
E(z,t) = x̂|Ex0
| cos(ω t − kz + φ + )
2π z
+ φ + (mV/m).
= x̂ 1.2π cos 2π × 106t −
300
(7.37)
(7.38)
where c is the velocity of light and η0 is called the intrinsic
impedance of free space.
(µ A/m),
where we have used the approximation η0 ≈ 120π (Ω).
At t = 0,
2π z π
E(z, 0) = x̂ 1.2π cos
−
(mV/m),
300 3
2π z π
−
(µ A/m).
H(z, 0) = ŷ 10 cos
300 3
Plots of E(z, 0) and H(z, 0) as a function of z are shown in
Fig. 7-5.
7-2
PLANE-WAVE PROPAGATION IN LOSSLESS MEDIA
307
Wave Propagating Along +z with E Along x̂
x
Let us apply Eq. (7.39a) to the wave given by Eq. (7.32a). The
e = x̂ Eex+ (z). Hence,
direction of propagation k̂ = ẑ and E
1.2π (mV/m)
Ee+ (z)
e = 1 k̂×
e = 1 (ẑ×
× x̂) Eex+ (z) = ŷ x
×E
H
,
η
η
η
λ
E
(7.40)
which is the same as the result given by Eq. (7.32b).
10 (μA/m) 0
y
H
Wave Propagating Along −z with E Along x̂
H
E
z
Figure 7-5 Spatial variations of E and H at t = 0 for the plane
wave of Example 7-1.
For a wave traveling in the −z direction with electric field
given by
e = x̂ Eex− (z) = x̂E − e jkz ,
(7.41)
E
x0
application of Eq. (7.39a) gives
E−
Ee− (z)
e = 1 (−ẑ×
× x̂) Eex− (z) = −ŷ x
= −ŷ x0 e jkz . (7.42)
H
η
η
η
e points in the negative y direction.
Hence, in this case, H
Wave Propagating Along +z with E Along x̂ and ŷ
7-2.2 General Relation between E and H
It can be shown that, for any uniform plane wave traveling in
an arbitrary direction denoted by the unit vector k̂, the electric
e and H
e are related as
and magnetic field phasors E
e = 1 k̂×
e
× E,
H
η
e
e = −η k̂×
× H.
E
In general, a uniform plane wave traveling in the +z direction
e is given by
may have both x and y components, where E
◮ The following right-hand rule applies: When we rotate
the four fingers of the right hand from the direction of E
toward that of H, the thumb points in the direction of the
wave travel, k̂. ◭
The relations given by Eqs. (7.39a and b) are valid not only
for lossless media but for lossy ones as well. As we see later in
Section 7-4, the expression for η of a lossy medium is different
from that given by Eq. (7.31). As long as the expression used
for η is appropriate for the medium in which the wave is
traveling, the relations given by Eqs. (7.39a and b) always
hold.
(7.43a)
e = x̂ H
ex+ (z) + ŷ H
ey+ (z).
H
(7.43b)
and the associated magnetic field is
(7.39a)
(7.39b)
e = x̂ Eex+ (z) + ŷ Eey+ (z),
E
Application of Eq. (7.39a) gives
Ee+ (z)
Ee+ (z)
e = 1 ẑ×
e = −x̂ y
×E
H
+ ŷ x
.
η
η
η
(7.44)
By equating Eq. (7.43b) to Eq. (7.44), we have
ex+ (z) = −
H
Eey+ (z)
,
η
e+
ey+ (z) = Ex (z) .
H
η
(7.45)
These results are illustrated in Fig. 7-6. The wave may be
considered the sum of two waves: one with electric and
magnetic components (Ex+ , Hy+ ) and another with components
(Ey+ , Hx+ ). In general, a TEM wave may have an electric field
in any direction in the plane orthogonal to the direction of wave
travel, the associated magnetic field is also in the same plane,
and its direction is dictated by Eq. (7.39a).
308
CHAPTER 7 PLANE-WAVE PROPAGATION
Module 7.1 Linking E to H Select the directions and magnitudes of E and H and observe the resultant wave vector.
What is a uniform plane wave?
Describe its properties, both physically and mathematically. Under what conditions is it appropriate to treat a
spherical wave as a plane wave?
Concept Question 7-1:
y
E
Ey+
H
e and H
e are governed by
Concept Question 7-2: Since E
wave equations of the same form [Eqs. (7.15) and (7.16)],
e = H?
e Explain.
does it follow that E
Hy+
If a TEM wave is traveling in
the ŷ direction, can its electric field have components
along x̂, ŷ, and ẑ? Explain.
Concept Question 7-3:
Hx+
z
Ex+
x
Figure 7-6 The wave (E, H) is equivalent to the sum of two
waves: one with fields (Ex+ , Hy+ ) and another with (Ey+ , Hx+ )
with both traveling in the +z direction.
Exercise 7-1: A 10-MHz uniform plane wave is traveling
in a nonmagnetic medium with µ = µ0 and εr = 9. Find
(a) the phase velocity, (b) the wavenumber, (c) the wavelength in the medium, and (d) the intrinsic impedance of
the medium.
7-3 WAVE POLARIZATION
309
Answer: (a) up = 1 × 108 m/s, (b) k = 0.2π rad/m, (c)
λ = 10 m, (d) η = 125.67 Ω. (See
EM
.)
Exercise 7-2: The electric field phasor of a uniform plane
wave traveling in a lossless medium with an intrinsic
e = ẑ 10e− j4π y (mV/m).
impedance of 188.5 Ω is given by E
Determine (a) the associated magnetic field phasor and (b)
the instantaneous expression for E(y,t) if the medium is
nonmagnetic (µ = µ0 ).
e = x̂ 53e− j4π y (µ A/m),
Answer: (a) H
(b) E(y,t) = ẑ 10 cos(6π × 108t − 4π y) (mV/m). (See
EM
.)
Exercise 7-3: If the magnetic field phasor of a plane
wave traveling in a medium with intrinsic impedance
e = (ŷ 10 + ẑ 20)e− j4x (mA/m),
η = 100 Ω is given by H
find the associated electric field phasor.
e = (−ẑ + ŷ2)e− j4x (V/m). (See
Answer: E
EM
.)
Exercise 7-4: Repeat Exercise 7-3 for a magnetic field
e = ŷ(10e− j3x − 20e j3x) (mA/m).
given by H
e = −ẑ(e− j3x + 2e j3x ) (V/m). (See
Answer: E
EM
.)
7-3 Wave Polarization
◮ The polarization of a uniform plane wave describes
the locus traced by the tip of the E vector (in the plane
orthogonal to the direction of propagation) at a given point
in space as a function of time. ◭
In the most general case, the locus of the tip of E is an ellipse,
and the wave is said to be elliptically polarized. Under certain
conditions, the ellipse may degenerate into a circle or a straight
line, in which case the polarization state is called circular or
linear, respectively.
It was shown in Section 7-2 that the z components of the
electric and magnetic fields of a z-propagating plane wave are
both zero. Hence, in the most general case, the electric field
e of a +z-propagating plane wave may consist of an
phasor E(z)
x component, x̂ Eex (z), and a y component, ŷ Eey (z), or
with
e = x̂Eex (z) + ŷEey (z)
E(z)
(7.46)
Eex (z) = Ex0 e− jkz ,
(7.47a)
Eey (z) = Ey0 e− jkz ,
(7.47b)
where Ex0 and Ey0 are the amplitudes of Eex (z) and Eey (z), respectively. For the sake of simplicity, the plus sign superscript
has been suppressed; the negative sign in e− jkz is sufficient to
remind us that the wave is traveling in the positive z direction.
The two amplitudes Ex0 and Ey0 are, in general, complex
quantities with each characterized by a magnitude and a phase
angle. The phase of a wave is defined relative to a reference
state, such as z = 0 and t = 0 or any other combination of z
and t. As will become clear from the discussion that follows,
the polarization of the wave described by Eqs. (7.46) and
(7.47) depends on the phase of Ey0 relative to that of Ex0 but not
on the absolute phases of Ex0 and Ey0 . Hence, for convenience,
we assign Ex0 a phase of zero and denote the phase of Ey0 ,
relative to that of Ex0 , as δ . Thus, δ is the phase difference
e Accordingly, we define
between the y and x components of E.
Ex0 and Ey0 as
Ex0 = ax ,
(7.48a)
Ey0 = ay e jδ ,
(7.48b)
where ax = |Ex0 | ≥ 0 and ay = |Ey0 | ≥ 0 are the magnitudes
of Ex0 and Ey0 , respectively. Thus, by definition, ax and ay
may not assume negative values. Using Eqs. (7.48a and b) in
Eqs. (7.47a and b), the total electric field phasor is
e = (x̂ax + ŷay e jδ )e− jkz ,
E(z)
(7.49)
= x̂ax cos(ω t − kz) + ŷay cos(ω t − kz + δ ).
(7.50)
and the corresponding instantaneous field is
h
i
e e jω t
E(z,t) = Re E(z)
When characterizing an electric field at a given point in space,
two of its attributes that are of particular interest are its
magnitude and direction. The magnitude of E(z,t) is
|E(z,t)| = [Ex2 (z,t) + Ey2(z,t)]1/2
= [a2x cos2 (ω t − kz) + a2y cos2 (ω t − kz + δ )]1/2.
(7.51)
The electric field E(z,t) has components along the x and y
directions. At a specific position z, the direction of E(z,t) is
characterized by its inclination angle ψ , defined with respect
to the x axis and given by
Ey (z,t)
.
(7.52)
ψ (z,t) = tan−1
Ex (z,t)
In the general case, both the intensity of E(z,t) and its direction
are functions of z and t. Next, we examine some special cases.
310
CHAPTER 7 PLANE-WAVE PROPAGATION
Module 7.2 Plane Wave Observe a plane wave propagating along the z direction; note the temporal and spatial variations
of E and H. Examine how the wave properties change as a function of the values selected for the wave parameters—frequency
and E field amplitude and phase—and the medium’s constitutive parameters (ε , µ , σ ).
7-3.1 Linear Polarization
◮ A wave is said to be linearly polarized if for a fixed z,
the tip of E(z,t) traces a straight line segment as a function
of time. This happens when Ex (z,t) and Ey (z,t) are in
phase (i.e., δ = 0) or out of phase (δ = π ). ◭
Under these conditions, Eq. (7.50) simplifies to
E(0,t) = (x̂ax + ŷay ) cos(ω t − kz),
E(0,t) = (x̂ax − ŷay ) cos(ω t − kz).
(in phase)
(7.53a)
(out of phase) (7.53b)
Let us examine the out-of-phase case. The field’s magnitude is
|E(z,t)| =
[a2x + a2y ]1/2 | cos(ω t − kz)|,
and the inclination angle is
−ay
.
ψ = tan−1
ax
(out of phase)
The length of the vector representing E(0,t) decreases to zero
at ω t = π /2. The vector then reverses direction and increases
in magnitude to [a2x + a2y ]1/2 in the second quadrant of the x–y
plane at ω t = π . Since ψ is independent of both z and t, E(z,t)
maintains a direction along the line making an angle ψ with
the x axis while oscillating back and forth across the origin.
If ay = 0, then ψ = 0◦ or 180◦, and the wave is x-polarized;
conversely, if ax = 0, then ψ = 90◦ or −90◦ , and the wave is
y-polarized.
(7.54a)
7-3.2
Circular Polarization
We now consider the special case when the magnitudes of the
e are equal, and the phase difference
x and y components of E(z)
δ = ±π /2. For reasons that become evident shortly, the wave
polarization is called left-hand circular when δ = π /2, and
right-hand circular when δ = −π /2.
(7.54b)
We note that ψ is independent of both z and t. Figure 7-7
displays the line segment traced by the tip of E at z = 0 over
a half of a cycle. The trace would be the same at any other
value of z as well. At z = 0 and t = 0, |E(0, 0)| = [a2x + a2y ]1/2 .
Left-Hand Circular (LHC) Polarization
For ax = ay = a and δ = π /2, Eqs. (7.49) and (7.50) become
e = (x̂a + ŷae jπ /2)e− jkz = a(x̂ + jŷ)e− jkz ,
E(z)
(7.55a)
7-3 WAVE POLARIZATION
311
y
z
z
y
E
a
x
E
ψ ω
(a) LHC polarization
Ey
ay
ωt = π
−ay
a
ωt = 0
ω
E
ψ
Figure 7-7
Linearly polarized wave traveling in the +z
direction (out of the page).
h
i
e e jω t
E(z,t) = Re E(z)
= x̂a cos(ω t − kz) + ŷa cos(ω t − kz + π /2)
= x̂a cos(ω t − kz) − ŷa sin(ω t − kz).
Figure 7-8 Circularly polarized plane waves propagating in
= [a2 cos2 (ω t − kz) + a2 sin2 (ω t − kz)]1/2 = a
(7.56a)
and
−1
= tan
Ey (z,t)
Ex (z,t)
−a sin(ω t − kz)
= −(ω t − kz).
a cos(ω t − kz)
x
(7.55b)
1/2
|E(z,t)| = Ex2 (z,t) + Ey2(z,t)
a
z
(b) RHC polarization
The corresponding field magnitude and inclination angle are
ψ (z,t) = tan−1
z
y
ax
Ex
ψ
−ax
x
a
z
the +z direction (out of the page).
the x–y plane and rotates in a clockwise direction as a function
of time (when viewing the wave approaching). Such a wave is
called left-hand circularly polarized because when the thumb
of the left hand points along the direction of propagation (the
z direction in this case) the other four fingers point in the
direction of rotation of E.
Right-Hand Circular (RHC) Polarization
(7.56b)
We observe that the magnitude of E is independent of both z
and t, whereas ψ depends on both variables. These functional
dependencies are the converse of those for the linear polarization case.
At z = 0, Eq. (7.56b) gives ψ = −ω t; the negative sign
implies that the inclination angle decreases as time increases.
As illustrated in Fig. 7-8(a), the tip of E(t) traces a circle in
For ax = ay = a and δ = −π /2, we have
|E(z,t)| = a,
ψ = (ω t − kz).
(7.57)
The trace of E(0,t) as a function of t is shown in Fig. 7-8(b).
For RHC polarization, the fingers of the right hand point
in the direction of rotation of E when the thumb is along
the propagation direction. Figure 7-9 depicts a right-hand
circularly polarized wave radiated by a helical antenna.
312
CHAPTER 7 PLANE-WAVE PROPAGATION
x
Transmitting
antenna
ω
E
E
y
y
Left screw sense
in space
x
z
z
Right sense of rotation
in plane
Figure 7-10 Right-hand circularly polarized wave of Example
7-2.
Figure 7-9 Right-hand circularly polarized wave radiated by a
helical antenna.
and application of Eq. (7.39a) gives
◮ Polarization handedness is defined in terms of the
rotation of E as a function of time in a fixed plane orthogonal to the direction of propagation, which is opposite the
direction of rotation of E as a function of distance at a
fixed point in time. ◭
1
e
e = 1 ŷ×
× E(y)
× (−x̂ j + ẑ)3e− jky
H(y)
= ŷ×
η
η
3
= (ẑ j + x̂)e− jky
(mA/m).
η
With ω = 2π f = 2π × 108 (rad/s), the wavenumber k is
√
√
ω εr 2π × 108 4 4
(rad/m),
= π
=
k=
c
3 × 108
3
and the intrinsic impedance η is
Example 7-2:
RHC-Polarized Wave
An RHC-polarized plane wave with electric field magnitude
of 3 (mV/m) is traveling in the +y direction in a dielectric
medium with ε = 4ε0 , µ = µ0 , and σ = 0. If the frequency is
100 MHz, obtain expressions for E(y,t) and H(y,t).
Solution: Since the wave is traveling in the +y direction, its
field must have components along the x and z directions. The
rotation of E(y,t) is depicted in Fig. 7-10, where ŷ is out of the
page. By comparison with the RHC-polarized wave shown in
e
Fig. 7-8(b), we assign the z component of E(y)
a phase angle
of zero and the x component a phase shift of δ = −π /2. Both
components have magnitudes of a = 3 (mV/m). Hence,
e
E(y)
= x̂Eex + ẑEez = x̂ae− jπ /2e− jky + ẑae− jky
= (−x̂ j + ẑ)3e− jky
(mV/m),
η0
120π
η = √ ≈ √ = 60π
εr
4
(Ω).
The instantaneous fields E(y,t) and H(y,t) are
h
i
e
E(y,t) = Re E(y)
e jω t
i
h
= Re (−x̂ j + ẑ)3e− jky e jω t
= 3[x̂ sin(ω t − ky) + ẑcos(ω t − ky)]
(mV/m)
and
h
i
e
H(y,t) = Re H(y)
e jω t
3
− jky j ω t
(ẑ j + x̂)e
e
= Re
η
=
1
[x̂ cos(ω t − ky) − ẑsin(ω t − ky)] (mA/m).
20π
7-3 WAVE POLARIZATION
313
Technology Brief 13: RFID Systems
In 1973, two separate patents were issued in the United
States for Radio Frequency Identification (RFID) concepts. The first, granted to Mario Cardullo, was for an
active RFID tag with rewritable memory. An active tag
has a power source (such as a battery) of its own,
whereas a passive RFID tag does not. The second
patent was granted to Charles Walton, who proposed
the use of a passive tag for keyless entry (unlocking a
door without a key). Shortly thereafter, a passive RFID
tag was developed for tracking cattle (Fig. TF13-1), and
the technology rapidly expanded into many commercial
enterprises—from tracking vehicles and consumer products to supply chain management and automobile antitheft systems.
Figure TF13-1 Passive RFID tags were developed in the
1970s for tracking cows.
RFID System Overview
In an RFID system, communication occurs between a
reader—which actually is a transceiver—and a tag
(Fig. TF13-2). When interrogated by the reader, a tag
responds with information about its identity, as well as
other relevant information depending on the specific
application.
◮ The tag is, in essence, a transponder commanded by the reader. ◭
The functionality and associated capabilities of the RFID
tag depend on two important attributes: (a) whether
the tag is of the active or passive type and (b) the
tag’s operating frequency. Usually the RFID tag remains
dormant (asleep) until activated by an electromagnetic
signal radiated by the reader’s antenna. The magnetic
field of the EM signal induces a current in the coil
contained in the tag’s circuit (Fig. TF13-3). For a passive
tag, the induced current has to be sufficient to generate
the power necessary to activate the chip as well as to
transmit the response to the reader.
◮ Passive RFID systems are limited to short read
ranges (between reader and tag) on the order of
30 cm to 3 m, depending on the system’s frequency
band (as noted in Table TT13-1). ◭
The obvious advantage of active RFID systems is that
they can operate over greater distances and do not
require reception of a signal from the reader’s antenna to
get activated. However, active tags are significantly more
expensive to fabricate than their passive cousins.
RFID Frequency Bands
Table TT13-1 provides a comparison among the four
frequency bands commonly used for RFID systems.
Generally speaking, the higher-frequency tags can operate over longer read ranges and can carry higher data
rates, but they are more expensive to fabricate.
314
CHAPTER 7 PLANE-WAVE PROPAGATION
Tag reader
The reader forwards the
data it received from the
RFID tag to a database
that can then match the
tag’s identifying serial
number to an authorized
account and debit that
account.
Once activated by the signal from the
tag reader (which acts as both a
transmitter and a receiver), the RFID
tag responds by transmitting the
identifying serial number programmed
into its electronic chip.
Figure TF13-2 How an RFID system works is illustrated through this EZ-Pass example. The image of the UHF RFID tag is courtesy of
Cary Wolinsky/Cavan Images/Alamy Stock Photo.
RFID reader
Chip
Antenna
Antenna
Tag
Figure TF13-3 Simplified diagram for how the RFID reader communicates with the tag. At the two lower carrier frequencies commonly
used for RFID communication (125 kHz and 13.56 MHz), coil inductors act as magnetic antennas. In systems designed to operate at higher
frequencies (900 MHz and 2.54 GHz), dipole antennas are used instead.
Table TT13-1 Comparison of RFID frequency bands.
Band
LF
HF
UHF
Microwave
RFID frequency
125–134 kHz
13.56 MHz
865–956 MHz
2.45 GHz
Read range
≤ 0.5 m
≤ 1.5 m
≤5m
≤ 10 m
Data rate
1 kbit/s
25 kbit/s
30 kbit/s
100 kbit/s
Typical
applications
• Animal ID
• Automobile key/antitheft
• Access control
• Smart cards
• Article surveillance
• Airline baggage tracking
• Library book tracking
• Supply chain
management
• Logistics
• Vehicle toll collection
• Railroad car monitoring
7-3 WAVE POLARIZATION
315
7-3.3 Elliptical Polarization
Plane waves that are not linearly or circularly polarized are
elliptically polarized. That is, the tip of E(z,t) traces an ellipse
in the plane perpendicular to the direction of propagation. The
shape of the ellipse and the field’s handedness (left-hand or
right-hand) are determined by the values of the ratio (ay /ax )
and the phase difference δ .
The polarization ellipse shown in Fig. 7-11 has its major
axis with length aξ along the ξ direction and its minor axis
with length aη along the η direction. The rotation angle γ is
defined as the angle between the major axis of the ellipse and
a reference direction, which is chosen here to be the x axis,
with γ being bounded within the range −π /2 ≤ γ ≤ π /2. The
shape of the ellipse and its handedness are characterized by the
ellipticity angle χ , which is defined as
tan χ = ±
aη
1
=± ,
aξ
R
(7.58)
with the plus sign corresponding to left-handed rotation and
the minus sign corresponding to right-handed rotation. The
limits for χ are −π /4 ≤ χ ≤ π /4. The quantity R = aξ /aη
is called the axial ratio of the polarization ellipse, and it
varies between 1 for circular polarization and ∞ for linear
polarization. The polarization angles γ and χ are related to the
wave parameters ax , ay , and δ by∗
tan 2γ = (tan 2ψ0 ) cos δ
(−π /2 ≤ γ ≤ π /2),
sin 2χ = (sin 2ψ0 ) sin δ
(7.59a)
(−π /4 ≤ χ ≤ π /4),
η
aξ
aη
ψ0
z
Major
axis
Minor
axis
x
Rotation
angle
Polarization ellipse
Figure 7-11 Polarization ellipse in the x–y plane with the wave
traveling in the z direction (out of the page).
Since the magnitudes ax and ay are, by definition, nonnegative numbers, the ratio ay /ax may vary between zero for an
x-polarized linear polarization and ∞ for a y-polarized linear
polarization. Consequently, the angle ψ0 is limited to the
range 0 ≤ ψ0 ≤ 90◦ . Application of Eq. (7.59a) leads to two
possible solutions for the value of γ , both of which fall within
the defined range from −π /2 to π /2. The correct choice is
governed by the following rule:
γ > 0 if cos δ > 0,
γ < 0 if cos δ < 0.
◮ In summary, the sign of the rotation angle γ is the same
as the sign of cos δ , and the sign of the ellipticity angle χ
is the same as the sign of sin δ . ◭
Example 7-3: Polarization State
− ŷ 4 sin(ω t − kz + 45◦)
γ
ax
◮ Positive values of χ , corresponding to sin δ > 0, are
associated with left-handed rotation, and negative values
of χ , corresponding to sin δ < 0, are associated with righthanded rotation. ◭
E(z,t) = x̂ 3 cos(ω t − kz + 30◦)
ξ
χ
Sketches of the polarization ellipse are shown in Fig. 7-12 for
various combinations of the angles (γ , χ ). The ellipse reduces
to a circle for χ = ±45◦ and to a line for χ = 0.
Determine the polarization state of a plane wave with electric
field
Ellipticity angle
ay
(7.60)
(7.59b)
∗ From M. Born and E. Wolf, Principles of Optics, New York: Macmillan,
1965, p. 27.
y
where ψ0 is an auxiliary angle defined by
ay
π
.
tan ψ0 =
0 ≤ ψ0 ≤
ax
2
(mV/m).
Solution: We begin by converting the second term to a cosine
reference:
E = x̂ 3 cos(ω t − kz + 30◦) − ŷ 4 cos(ω t − kz + 45◦ − 90◦)
= x̂ 3 cos(ω t − kz + 30◦) − ŷ 4 cos(ω t − kz − 45◦).
e is
The corresponding field phasor E(z)
e = x̂ 3e− jkz e j30◦ − ŷ 4e− jkz e− j45◦
E(z)
◦
◦
= x̂ 3e− jkz e j30 + ŷ 4e− jkz e− j45 e j180
◦
◦
= x̂ 3e− jkz e j30 + ŷ 4e− jkz e j135 ,
◦
316
CHAPTER 7 PLANE-WAVE PROPAGATION
χ
γ
45◦
22.5◦
0◦
−22.5◦
−45◦
−90◦
−45◦
0◦
45◦
90◦
Left circular polarization
Left elliptical polarization
Linear polarization
Right elliptical polarization
Right circular polarization
Figure 7-12 Polarization states for various combinations of the polarization angles (γ , χ ) for a wave traveling out of the page.
where we have replaced the negative sign of the second term
◦
with e j180 in order to have positive amplitudes for both
terms, thereby allowing us to use the definitions given earlier.
e
According to the expression for E(z),
the phase angles of the x
and y components are δx = 30◦ and δy = 135◦ , giving a phase
difference of δ = δy − δx = 135◦ − 30◦ = 105◦ . The auxiliary
angle ψ0 is obtained from
ay
4
ψ0 = tan−1
= 53.1◦.
= tan−1
ax
3
From Eq. (7.59a),
tan 2γ = (tan 2ψ0 ) cos δ = tan 106.2◦ cos 105◦ = 0.89,
which gives two solutions for γ , namely γ = 20.8◦ and
γ = −69.2◦. Since cos δ < 0, the correct value of γ is −69.2◦.
From Eq. (7.59b),
Which of the following two
descriptions defines an RHC-polarized wave: A wave
incident upon an observer is RHC-polarized if its electric
field appears to the observer to rotate in a counterclockwise direction (a) as a function of time in a fixed plane
perpendicular to the direction of wave travel or (b) as a
function of travel distance at a fixed time t?
Concept Question 7-5:
Exercise 7-5: The electric field of a plane wave is given
by
E(z,t) = x̂ 3 cos(ω t − kz) + ŷ4 cos(ω t − kz) (V/m).
sin 2χ = (sin 2ψ0 ) sin δ
= sin 106.2◦ sin 105◦ = 0.93 or
An elliptically polarized wave
is characterized by amplitudes ax and ay and by the phase
difference δ . If ax and ay are both nonzero, what should
δ be in order for the polarization state to reduce to linear
polarization?
Concept Question 7-4:
χ = 34.0◦.
The magnitude of χ indicates that the wave is elliptically
polarized, and its positive polarity specifies its rotation as left
handed.
Determine (a) the polarization state, (b) the modulus of E,
and (c) the auxiliary angle.
(a) Linear, (b) |E| = 5 cos(ω t − kz) (V/m),
(c) ψ0 = 53.1◦. (See EM .)
Answer:
7-4
PLANE-WAVE PROPAGATION IN LOSSY MEDIA
317
Module 7.3 Polarization I Upon specifying the amplitudes and phases of the x and y components of E, the user can
observe the trace of E in the x–y plane.
Exercise 7-6: If the electric field phasor of a TEM wave
e = (ŷ − ẑ j)e− jkx , determine the polarization
is given by E
state.
Answer: RHC polarization. (See
EM
.)
7-4 Plane-Wave Propagation in Lossy
Media
where ε ′ = ε and ε ′′ = σ /ω . Since γ is complex, we express
it as
γ = α + jβ ,
(7.63)
where α is the medium’s attenuation constant and β its phase
constant. By replacing γ with (α + jβ ) in Eq. (7.62), we
obtain
(α + jβ )2 = (α 2 − β 2 )+ j2αβ = −ω 2 µε ′ + jω 2 µε ′′ . (7.64)
To examine wave propagation in a lossy (conducting) medium,
we return to the wave equation given by Eq. (7.15),
The rules of complex algebra require the real and imaginary
parts on one side of an equation to equal, respectively, the real
and imaginary parts on the other side. Hence,
e=0
e − γ 2E
∇2 E
α 2 − β 2 = −ω 2 µε ′ ,
with
γ 2 = −ω 2 µεc = −ω 2 µ (ε ′ − jε ′′ ),
(7.61)
(7.62)
2αβ = ω 2 µε ′′ .
(7.65a)
(7.65b)
318
CHAPTER 7 PLANE-WAVE PROPAGATION
Module 7.4 Polarization II Upon specifying the amplitudes and phases of the x and y components of E, the user can
observe the 3-D profile of the E vector over a specified length span.
Solving these two equations for α and β gives

s
1/2
′′ 2
 µε ′

 1+ ε
− 1
α =ω
′
 2

ε
(Np/m),

s
1/2
′′ 2
 µε ′

ε
 1+

β =ω
+
1
 2

ε′
(7.66a)
e
e = ŷ H
ey (z) = ŷ Ex (z) = ŷ Ex0 e−α z e− jβ z ,
H(z)
ηc
ηc
ηc =
e = x̂ Eex (z)
For a uniform plane wave with electric field E
traveling along the z direction, the wave equation given by
Eq. (7.61) reduces to
(7.67)
The general solution of the wave equation given by Eq. (7.67)
comprises two waves: one traveling in the +z direction and
another traveling in the −z direction. Assuming only the
former is present, the solution of the wave equation leads to
e = x̂Eex (z) = x̂Ex0 e−γ z = x̂Ex0 e−α z e− jβ z .
E(z)
(7.69)
where
(7.66b)
(rad/m).
d 2 Eex (z)
− γ 2 Eex (z) = 0.
dz2
e can be determined by applyThe associated magnetic field H
e = − jω µ H,
e or using Eq. (7.39a):
ing Eq. (7.2b): ∇ × E
e
e
×
H = (k̂ E)/ηc , where ηc is the intrinsic impedance of the
lossy medium. Both approaches give
(7.68)
r
µ
=
εc
r
µ
ε′
ε ′′ −1/2
1− j ′
ε
(Ω).
(7.70)
We noted earlier that in a lossless medium, E(z,t) is in phase
with H(z,t). This property no longer holds true in a lossy
medium because ηc is complex. This fact is demonstrated later
in Example 7-4.
From Eq. (7.68), the magnitude of Eex (z) is given by
|Eex (z)| = |Ex0 e−α z e− jβ z| = |Ex0 |e−α z ,
(7.71)
which decreases exponentially with z at a rate dictated by the
ey = Eex /ηc , the magnitude of H
ey
attenuation constant α . Since H
α
z
−
also decreases as e . As the field attenuates, part of the
energy carried by the electromagnetic wave is converted into
7-4
PLANE-WAVE PROPAGATION IN LOSSY MEDIA
1
319
a low-loss dielectric, and when ε ′′ /ε ′ ≫ 1, it is considered a
good conductor. In practice, the medium may be regarded as
a low-loss dielectric if ε ′′ /ε ′ < 10−2 , as a good conductor if
ε ′′ /ε ′ > 102 , and as a quasi-conductor if 10−2 ≤ ε ′′ /ε ′ ≤ 102 .
For low-loss dielectrics and good conductors, the expressions
given by Eq. (7.66) can be significantly simplified, as shown
next.
~
|Ex(z)|
|Ex0|
e−αz
e−1
7-4.1 Low-Loss Dielectric
z
δs
From Eq. (7.62), the general expression for γ is
Figure 7-13 Attenuation of the magnitude of Eex (z) with
distance z. The skin depth δs is the value of z at which
|Eex (z)|/|Ex0 | = e−1 or z = δs = 1/α .
heat due to conduction in the medium. As the wave travels
through a distance z = δs with
δs =
1
α
(m),
γ = jω
◮ This distance δs , called the skin depth of the medium,
characterizes how deep an electromagnetic wave can penetrate into a conducting medium. ◭
In a perfect dielectric, σ = 0 and ε ′′ = 0. Use of Eq. (7.66a)
yields α = 0; therefore, δs = ∞. Thus, in free space, a plane
wave can propagate indefinitely with no loss in magnitude.
On the other extreme, in a perfect conductor, σ = ∞ and use
of Eq. (7.66a) leads to α = ∞. Hence δs = 0. If the outer
conductor of a coaxial cable is designed to be several skin
depths thick, it prevents energy inside the cable from leaking
outward and shields against penetration of electromagnetic
energy from external sources into the cable.
The expressions given by Eqs. (7.66a), (7.66b), and (7.70)
for α , β , and ηc are valid for any linear, isotropic, and
homogeneous medium. For a perfect dielectric (σ = 0), these
expressions reduce to those for√the lossless case (Section 7-2),
wherein α = 0, β = k = ω µε , and ηc = η . For a lossy
medium, the ratio ε ′′ /ε ′ = σ /ωε , which appears in all these
expressions, plays an important role in classifying how lossy
the medium is. When ε ′′ /ε ′ ≪ 1, the medium is considered
ε ′′ 1/2
µε ′ 1 − j ′
.
ε
(7.73)
For |x| ≪ 1, the function (1 − x)1/2 can be approximated by
the first two terms of its binomial series; that is, (1 − x)1/2 ≈
1 − x/2. By applying this approximation to Eq. (7.73) for a
low-loss dielectric with x = jε ′′ /ε ′ and ε ′′ /ε ′ ≪ 1, we obtain
(7.72)
the wave magnitude decreases by a factor of e−1 ≈ 0.37
(Fig. 7-13). At depth z = 3δs , the field magnitude is less than
5% of its initial value, and at z = 5δs , it is less than 1%.
p
γ ≈ jω
p
µε ′
ε ′′
1− j ′ .
2ε
(7.74)
The real and imaginary parts of Eq. (7.74) are
r
r
ωε ′′ µ
σ µ
α≈
=
2
ε′
2 ε
p
√
′
β ≈ ω µε = ω µε
(Np/m),
(rad/m).
(7.75a)
(7.75b)
(low-loss medium)
We note that the expression for β is the same as that for the
wavenumber k of a lossless medium. Applying the binomial
approximation (1 − x)−1/2 ≈ (1 + x/2) to Eq. (7.70) leads to
ηc ≈
r
r µ
ε ′′
µ
σ 1
+
j
=
1
+
j
.
ε′
2ε ′
ε
2ωε
(7.76a)
In practice, because ε ′′/ε ′ = σ /ωε < 10−2 , the second term in
Eq. (7.76a) often is ignored. Thus,
ηc ≈
r
µ
,
ε
which is the same as Eq. (7.31) for the lossless case.
(7.76b)
320
Technology Brief 14:
Liquid Crystal Display (LCD)
LCDs are used in digital clocks, cellular phones, desktop and laptop computers, some televisions, and other
electronic systems. They offer a decided advantage over
former display technologies, such as cathode ray tubes,
because they are much lighter and thinner and consume
a lot less power to operate. LCD technology relies
on special electrical and optical properties of a class
of materials known as liquid crystals, which were first
discovered in the 1880s by botanist Friedrich Reinitzer.
Physical Principle
◮ Liquid crystals are neither a pure solid nor a pure
liquid; rather, they are a hybrid of both. ◭
One particular variety of interest is the twisted nematic
liquid crystal whose rod-shaped molecules have a
natural tendency to assume a twisted spiral structure when the material is sandwiched between finely
grooved glass substrates with orthogonal orientations
(Fig. TF14-1). Note that the molecules in contact with
the grooved surfaces align themselves in parallel along
the grooves, from a y orientation at the entrance substrate into an x orientation at the exit substrate. The
molecular spiral causes the crystal to behave like a wave
polarizer: unpolarized light incident upon the entrance
substrate follows the orientation of the spiral, emerging
through the exit substrate with its polarization (direction
of electric field) parallel to the groove’s direction, which
in Fig. TF14-1 is along the x direction. Thus, of the x and
y components of the incident light, only the y component
is allowed to pass through the y-polarized filter, but as a
consequence of the spiral action facilitated by the liquid
crystal’s molecules, the light that emerges from the LCD
structure is x-polarized.
CHAPTER 7 PLANE-WAVE PROPAGATION
x-polarized light
x
x-oriented
polarizing filter
x-oriented
exit substrate
Rod-shaped
molecules
Orthogonal
groove
orientations
y-oriented
entrance substrate
y-oriented
polarizing filter
x-polarized
component of
incident light
Only y-polarized
component can
pass through
x
y polarizing filter
Unpolarized light
Figure TF14-1 The rod-shaped molecules of a liquid crystal
sandwiched between grooved substrates with orthogonal orientations causes the electric field of the light passing through it to
rotate by 90◦ .
◮ The sandwiched liquid-crystal layer (typically on
the order of 5 microns in thickness or 1/20 of the
width of a human hair) is straddled by a pair of optical
filters with orthogonal polarizations. ◭
LCD Structure
A single-pixel LCD structure is shown in Fig. TF14-2
for the OFF and ON states, with OFF corresponding to a
bright-looking pixel and ON to a dark-looking pixel.
When no voltage is applied across the crystal layer
(Fig. TF14-2(a)), incoming unpolarized light gets polarized as it passes through the entrance polarizer, rotates
by 90◦ as it follows the molecular spiral, and finally
7-4
PLANE-WAVE PROPAGATION IN LOSSY MEDIA
Bright pixel
V
+
_
321
Dark pixel
Liquid crystal
V
5 μm
+
_
Molecule of
liquid crystal
Polarizing filter
(a) ON state (switch open)
(b) OFF state (switch closed)
Figure TF14-2 Single-pixel LCD.
emerges from the exit polarizer, giving the exited surface
a bright appearance. A useful feature of nematic liquid
crystals is that their spiral untwists (Fig. TF14-2(b))
under the influence of an electric field (induced by
a voltage difference across the layer). The degree of
untwisting depends on the strength of the electric field.
With no spiral to rotate the wave polarization as the
light travels through the crystal, the light polarization
becomes orthogonal to that of the exit polarizer, allowing
no light to pass through it. Hence, the pixel exhibits a
dark appearance.
◮ By extending the concept to a two-dimensional
array of pixels and devising a scheme to control
the voltage across each pixel individually (usually
by using a thin-film transistor), a complete image
can be displayed as illustrated in Fig. TF14-3. For
color displays (Fig. TF14-4), each pixel is made up
of three subpixels with complementary color filters
(red, green, and blue). ◭
322
CHAPTER 7 PLANE-WAVE PROPAGATION
2-D pixel array
Liquid crystal
678
Unpolarized light
Exit polarizer
Entrance polarizer
Molecular spiral
LCD display
Figure TF14-3 2-D LCD array.
Figure TF14-4 LCD display.
7-4
PLANE-WAVE PROPAGATION IN LOSSY MEDIA
7-4.2 Good Conductor
ε ′′ /ε ′
When
> 100, Eqs. (7.66a), (7.66b), and (7.70) can be
approximated as
r
r
µε ′′
µσ p
α ≈ω
= π f µσ (Np/m),
=ω
2
2ω
p
β ≈ α ≈ π f µσ (rad/m),
r
r
µ
π fµ
α
ηc ≈ j ′′ = (1 + j)
= (1 + j)
(Ω).
ε
σ
σ
323
This qualifies seawater as a good conductor at 1 kHz and
allows us to use the good-conductor expressions given in
Table 7-1:
p
π f µσ
p
= π × 103 × 4π × 10−7 × 4
α=
(7.77a)
= 0.126
(7.77b)
(good conductors)
Example 7-4:
Plane Wave in Seawater
A uniform plane wave is traveling in seawater. Assume that
the x–y plane resides just below the sea surface and the wave
travels in the +z direction into the water. The constitutive
parameters of seawater are εr = 80, µr = 1, and σ = 4 S/m. If
the magnetic field at z = 0 is
H(0,t) = ŷ 100 cos(2π × 103t + 15◦)
(mA/m),
(a) obtain expressions for E(z,t) and H(z,t), and
(b) determine the depth at which the magnitude of E is 1% of
its value at z = 0.
Solution: (a) Since H is along ŷ and the propagation direction
is ẑ, E must be along x̂. Hence, the general expressions for the
phasor fields are
e = x̂Ex0 e−α z e− jβ z ,
E(z)
e = ŷ Ex0 e−α z e− jβ z .
H(z)
ηc
(7.78a)
ε ′′
σ
σ
4
= 9 × 105.
=
=
=
ε′
ωε ωεr ε0 2π × 103 × 80 × (10−9/36π )
(7.79b)
(Ω).
(7.79c)
As no explicit information has been given about the electric
field amplitude Ex0 , we should assume it to be complex;
that is, Ex0 = |Ex0 |e jφ0 . The wave’s instantaneous electric and
magnetic fields are given by
i
h
E(z,t) = Re x̂|Ex0 |e jφ0 e−α z e− jβ z e jω t
= x̂|Ex0 |e−0.126z cos(2π × 103t − 0.126z + φ0) (V/m),
(7.80a)
|Ex0 |e jφ0 −α z − jβ z jω t
e
e e
H(z,t) = Re ŷ
0.044e jπ /4
= ŷ22.5|Ex0|e−0.126z cos(2π × 103t
− 0.126z + φ0 − 45◦) (A/m).
(7.80b)
At z = 0,
H(0,t) = ŷ 22.5|Ex0| cos(2π × 103t + φ0 − 45◦) (A/m).
(7.81)
By comparing Eq. (7.81) with the expression given in the
problem statement,
H(0,t) = ŷ 100 cos(2π × 103t + 15◦)
(mA/m),
we deduce that
(7.78b)
To determine α , β , and ηc for seawater, we begin by evaluating
the ratio ε ′′ /ε ′ . From the argument of the cosine function
of H(0,t), we deduce that ω = 2π × 103 (rad/s). Therefore,
f = 1 kHz. Hence,
(7.79a)
β = α = 0.126
(rad/m),
α
ηc = (1 + j)
σ
√ jπ /4 0.126
)
= 0.044e jπ /4
=( 2e
4
(7.77c)
In
√ we used the relation given by Eq. (1.53):
√ Eq. (7.77c),
j = (1 + j)/ 2. For a perfect conductor with σ = ∞, these
expressions yield α = β = ∞ and ηc = 0. A perfect conductor
is equivalent to a short circuit in a transmission line equivalent.
Expressions for the propagation parameters in various types
of media are summarized in Table 7-1.
(Np/m),
22.5|Ex0| = 100 × 10−3
or
|Ex0 | = 4.44
(mV/m)
and
φ0 − 45◦ = 15◦
or
φ0 = 60◦ .
324
CHAPTER 7 PLANE-WAVE PROPAGATION
Table 7-1 Expressions for α , β , ηc , up , and λ for various types of media.
Any Medium
α=
β=
ηc =
Lossless
Medium
(σ = 0)
1/2
s
′′ 2
′
µε
ε
 1+
− 1
ω
2
ε′

s
1/2
′′ 2
′
µε
ε
 1+
+ 1
ω
2
ε′

r
µ
ε′
1− j
ε ′′
ε′
−1/2
0
√
ω µε
r
µ
ε
up =
ω /β
√
1/ µε
λ=
2π /β = up / f
up / f
Low-loss
Medium
(ε ′′ /ε ′ ≪ 1)
r
Good
Conductor
(ε ′′ /ε ′ ≫ 1)
µ
ε
≈
p
π f µσ
(Np/m)
√
≈ ω µε
≈
p
π f µσ
(rad/m)
σ
≈
2
≈
r
µ
ε
√
≈ 1/ µε
≈ up / f
≈ (1 + j)
≈
p
α
σ
E(z,t) = x̂ 4.44e−0.126z cos(2π × 103t − 0.126z + 60◦)
(mV/m),
(7.82a)
H(z,t) = ŷ 100e−0.126z cos(2π × 103t − 0.126z + 15◦)
(mA/m).
(7.82b)
(Ω)
4π f /µσ
(m/s)
≈ up / f
Notes: ε ′ = ε ; ε ′′ = σ /ω ; in free space, ε = ε0 , µ = µ0 ; in practice, a material is considered
if ε ′′ /ε ′ = σ /ωε < 0.01 and a good conducting medium if ε ′′ /ε ′ > 100.
Hence, the final expressions for E(z,t) and H(z,t) are
Units
(m)
a low-loss medium
Exercise 7-7: The constitutive parameters of copper
are µ = µ0 = 4π × 10−7 (H/m), ε = ε0 ≈ (1/36π ) ×
10−9 (F/m), and σ = 5.8 × 107 (S/m). Assuming that
these parameters are frequency independent, over what
frequency range of the electromagnetic spectrum (see
Fig. 1-16) is copper a good conductor?
Answer: f < 1.04 × 1016 Hz, which includes the radio,
infrared, visible, and part of the ultraviolet regions of the
EM spectrum. (See EM .)
◮ Because E(z,t) and H(z,t) in a lossy medium no
longer have the same constant-phase angle, they no longer
oscillate in sync with one another as a function of z
and t. ◭
Exercise 7-8: Over what frequency range may dry soil
with εr = 3, µr = 1, and σ = 10−4 (S/m) be regarded as a
low-loss dielectric?
Answer: f > 60 MHz. (See
EM
.)
(b) The depth at which the amplitude of E has decreased to 1%
of its initial value at z = 0 is obtained from
0.01 = e−0.126z
Exercise 7-9: For a wave traveling in a medium with a
skin depth δs , what is the amplitude of E at a distance
of 3δs compared with its initial value?
Answer: e−3 ≈ 0.05 or 5%. (See
or
z=
ln(0.01)
= 36.55 m ≈ 37 m.
−0.126
EM
.)
7-5 CURRENT FLOW IN A GOOD CONDUCTOR
325
Module 7.5 Wave Attenuation Observe the profile of a plane wave propagating in a lossy medium. Determine the skin
depth, the propagation parameters, and the intrinsic impedance of the medium.
7-5 Current Flow in a Good Conductor
When a dc voltage source is connected across the ends of
a conducting wire, the current flowing through the wire is
uniformly distributed over its cross section. That is, the current
density J is the same along the axis of the wire and along
its outer perimeter (Fig. 7-14(a)). This is not true in the
ac case. As we see shortly, a time-varying current density
is maximum along the perimeter of the wire and decreases
exponentially as a function of distance toward the axis of the
wire (Fig. 7-14(b)). In fact, at very high frequencies, most of
the current flows in a thin layer near the wire surface, and if the
wire material is a perfect conductor, the current flows entirely
on the surface of the wire.
Before analyzing a wire with circular cross section, let us
consider the simpler geometry of a semi-infinite conducting
solid, as shown in Fig. 7-15(a). The solid’s planar interface
with a perfect dielectric is the x–y plane. If at z = 0− (just
e = x̂E0
above the surface) an x-polarized electric field with E
exists in the dielectric, a similarly polarized field is induced in
the conducting medium and propagates as a plane wave along
the +z direction. As a consequence of the boundary condition
mandating continuity of the tangential component of E across
the boundary between any two contiguous media, the electric
e
field at z = 0+ (just below the boundary) is E(0)
= x̂E0 also.
J
I
R
V
(a) dc case
J
I
R
V(t)
(b) ac case
Figure 7-14 Current density J in a conducting wire is (a)
uniform across its cross section in the dc case, but (b) in the
ac case, J is highest along the wire’s perimeter.
326
CHAPTER 7 PLANE-WAVE PROPAGATION
as expressed by Eq. (7.77b), Eq. (7.85) can be written as
E0
Jex (z) = J0 e−(1+ j)z/δs
J0
H0
x
Ie = w
~
(a) Exponentially decaying Jx(z)
l
∞
Z ∞
0
Jex (z) dz = w
Z ∞
0
J0 e−(1+ j)z/δs dz =
δs
z
(b) Equivalent J0 over skin depth δs
Figure 7-15 Exponential decay of current density Jex (z) with
z in a solid conductor. The total current flowing through (a)
a section of width w extending between z = 0 and z = ∞ is
equivalent to (b) a constant current density J0 flowing through
a section of depth δs .
Z=
1+ j l
Ve
=
e
σ δs w
I
The EM fields at any depth z in the conductor are given by
(7.83a)
e = ŷ E0 e−α z e− jβ z .
H(z)
ηc
with
Jex (z) = σ E0 e
e
= J0 e
−α z − j β z
e
,
(7.85)
where J0 = σ E0 is the amplitude of the current density at
the surface. In terms of the skin depth δs = 1/α defined by
Eq. (7.72) and using the fact that in a good conductor α = β
(Ω).
(7.89)
l
,
w
(7.90)
where Zs is called the internal or surface impedance of the
conductor and is defined as the impedance Z for a length
l = 1 m and a width w = 1 m. Thus,
(7.83b)
From J = σ E, the current flows in the x direction, and its
density is
e
J(z) = x̂ Jex (z),
(7.84)
−α z − j β z
(A).
It is customary to represent Z as
Z = Zs
e = x̂E0 e−α z e− jβ z ,
E(z)
J0 wδs
(1 + j)
(7.87)
The numerator of Eq. (7.87) is reminiscent of a uniform current
density J0 flowing through a thin surface of width w and
depth δs . Because Jex (z) decreases exponentially with depth z,
a conductor of finite thickness d can be considered electrically
equivalent to one of infinite depth as long as d exceeds a few
skin depths. Indeed, if d = 3δs [instead of ∞ in the integral of
Eq. (7.87)], the error incurred in using the result on the righthand side of Eq. (7.87) is less than 5%; if d = 5δs , the error is
less than 1%.
The voltage across a length l at the surface (Fig. 7-15(b)) is
given by
J0
Ve = E0 l =
l.
(7.88)
σ
Hence, the impedance of a slab of width w, length l, and depth
d = ∞ (or, in practice, d > 5δs ) is
z
w
(7.86)
The current flowing through a rectangular strip of width w
along the y direction and extending between zero and ∞ in the
z direction is
~
Jx(z)
J0
(A/m2 ).
Zs =
1+ j
σ δs
(Ω).
(7.91)
Since the reactive part of Zs is positive, Zs can be defined as
Zs = Rs + jω Ls
with
r
π fµ
(Ω),
σ
r
1
µ
1
Ls =
(H),
=
ωσ δs 2 π f σ
1
=
Rs =
σ δs
(7.92a)
(7.92b)
7-6 ELECTROMAGNETIC POWER DENSITY
327
Similarly, for the outer conductor, the current is concentrated
within a thin layer of depth δs on the inside surface of the
conductor adjacent to the insulating medium between the two
conductors, which is where the EM fields exist. The resistance
per unit length for the outer conductor with radius b is
Outer conductor
Dielectric
2a 2b
Dielectric
R′2 =
Inner conductor
(a) Coaxial cable
Rs
2π b
(Ω/m),
(7.95)
and the coaxial cable’s total ac resistance per unit length is
2πa
R
′
= R′1 + R′2
Rs
=
2π
δs
1 1
+
a b
(Ω/m).
(7.96)
This expression was used in Chapter 2 for characterizing the
resistance per unit length of a coaxial transmission line.
(b) Equivalent inner conductor
Figure 7-16 The inner conductor of the coaxial cable in (a)
is represented in (b) by a planar conductor of width 2π a and
depth δs , as if its skin has been cut along its length on the bottom
side and then unfurled into a planar geometry.
How does β of a low-loss dielectric medium compare to that of a lossless medium?
Concept Question 7-6:
In a good conductor, does the
phase of H lead or lag that of E and by how much?
Concept Question 7-7:
√
where we used the relation δs = 1/α ≈ 1/ π f µσ given by
Eq. (7.77a). In terms of the surface resistance Rs , the ac
resistance of a slab of width w and length l is
R = Rs
l
l
=
w σ δs w
(Ω).
Attenuation means that a wave
loses energy as it propagates in a lossy medium. What
happens to the lost energy?
Concept Question 7-8:
(7.93)
Is a conducting medium dispersive or dispersionless? Explain.
Concept Question 7-9:
The expression for the ac resistance R is equivalent to the dc
resistance of a plane conductor of length l and cross section
A = δs w.
The results obtained for the planar conductor can be
extended to the coaxial cable shown in Fig. 7-16(a). If the
7
conductors are made of copper
√ with σ = 5.8 × 10 S/m, the
skin depth at 1 √
MHz is δs = 1/ π f µσ = 0.066 mm, and since
δs varies as 1/ f , it becomes smaller at higher frequencies.
As long as the inner conductor’s radius a is greater than 5δs ,
or 0.33 mm at 1 MHz, its “depth” may be regarded as
infinite. A similar criterion applies to the thickness of the outer
conductor. To compute the resistance of the inner conductor,
note that the current is concentrated near its outer surface and
approximately equivalent to a uniform current flowing through
a thin layer of depth δs and circumference 2π a. In other words,
the inner conductor’s resistance is nearly the same as that of a
planar conductor of depth δs and width w = 2π a, as shown in
Fig. 7-16(b). The corresponding resistance per unit length is
obtained by setting w = 2π a and dividing by l in Eq. (7.93):
R′1 =
R
Rs
=
l
2π a
(Ω/m).
(7.94)
Compare the flow of current
through a wire in the dc and ac cases. Compare the
corresponding dc and ac resistances of the wire.
Concept Question 7-10:
7-6 Electromagnetic Power Density
This section deals with the flow of power carried by an
electromagnetic wave. For any wave with an electric field E
and magnetic field H, the Poynting vector S is defined as
×H
S = E×
(W/m2 ).
(7.97)
The unit of S is (V/m) × (A/m) = (W/m2 ), and the direction
of S is along the wave’s direction of propagation. Thus, S
represents the power per unit area (or power density) carried
by the wave. If the wave is incident upon an aperture of area
A with outward surface unit vector n̂ as shown in Fig. 7-17,
328
CHAPTER 7 PLANE-WAVE PROPAGATION
Module 7.6 Current in a Conductor Module displays exponential decay of current density in a conductor.
S
Since both E and H are functions of time, so is the Poynting
vector S. In practice, however, the quantity of greater interest
is the average power density of the wave, Sav , which is the
time-average value of S:
nˆ
θ
kˆ
A
Sav =
Figure 7-17 EM power flow through an aperture.
A
S · n̂dA
(W).
h
i
e×H
e∗
Re E×
(W/m2 ).
(7.100)
This expression may be regarded as the electromagnetic equivalent of Eq. (2.107) for the time-average power carried by a
transmission line, namely
then the total power that flows through or is intercepted by the
aperture is
Z
P=
1
2
Pav (z) =
(7.98)
For a uniform plane wave propagating in a direction k̂ that
makes an angle θ with n̂, P = SA cos θ , where S = |S|.
Except for the fact that the units of S are per unit area,
Eq. (7.97) is the vector analogue of the scalar expression for
the instantaneous power P(z,t) flowing through a transmission
line,
P(z,t) = v(z,t) i(z,t),
(7.99)
where v(z,t) and i(z,t) are the instantaneous voltage and
current on the line.
1
2
h
i
Re Ve (z) Ie∗ (z) ,
(7.101)
e are the phasors corresponding to v(z,t)
where Ve (z) and I(z)
and i(z,t), respectively.
7-6.1
Plane Wave in a Lossless Medium
Recall that the general expression for the electric field of a
uniform plane wave with arbitrary polarization traveling in the
+z direction is
e = x̂ Eex (z) + ŷ Eey (z) = (x̂ Ex0 + ŷ Ey0 )e− jkz ,
E(z)
(7.102)
7-6 ELECTROMAGNETIC POWER DENSITY
329
where, in the general case, Ex0 and Ey0 may be complex
e is
quantities. The magnitude of E
e = (E
e ·E
e ∗ )1/2 = [|Ex0 |2 + |Ey0 |2 ]1/2 .
|E|
S
(7.103)
S
S
e is obtained by
The phasor magnetic field associated with E
applying Eq. (7.39a):
e
ex + ŷ H
ey )e− jkz
H(z)
= (x̂ H
=
1
e = 1 (−x̂ Ey0 + ŷ Ex0 )e− jkz .
×E
ẑ×
η
η
(7.104)
The wave can be considered as the sum of two waves: one comey ) and another comprising fields (Eey , H
ex ).
prising fields (Eex , H
Use of Eqs. (7.102) and (7.104) in Eq. (7.100) leads to
Sav = ẑ
e2
1
|E|
(|Ex0 |2 + |Ey0 |2 ) = ẑ
2η
2η
(W/m2 ),
(7.105)
(lossless medium)
Sun
S
Rs
S
Earth
Area of
spherical surface
2
Asph = 4πRs
(a) Radiated solar power
which states that power flows in the z direction with average
power density equal to the sum of the average power densities
ey ) and (Eey , H
ex ) waves. Note that, because Sav
of the (Eex , H
e
depends only on η and |E|, waves characterized by different
polarizations carry the same amount of average power as long
as their electric fields have the same magnitudes.
Example 7-5:
Solar Power
If solar illumination is characterized by a power density of
1 kW/m2 on the Earth’s surface, find (a) the total power radiated by the sun, (b) the total power intercepted by the Earth,
and (c) the electric field of the power density incident upon the
Earth’s surface, assuming that all the solar illumination is at
a single frequency. The radius of the Earth’s orbit around the
sun, Rs , is approximately 1.5 × 108 km, and the Earth’s mean
radius Re is 6,380 km.
Solution: (a) Assuming that the sun radiates isotropically
(equally in all directions), the total power it radiates is Sav Asph ,
where Asph is the area of a spherical shell of radius Rs
(Fig. 7-18(a)). Thus,
Psun = Sav (4π R2s ) = 1 × 103 × 4π × (1.5 × 1011)2
= 2.8 × 1026 W.
(b) With reference to Fig. 7-18(b), the power intercepted by
the Earth’s cross section Ae = π R2e is
Pint = Sav (π R2e ) = 1×103 × π ×(6.38×106)2 = 1.28×1017 W.
S
Sun
S
Ae = πRe2
Earth
(b) Earth intercepted power
Figure 7-18 Solar radiation intercepted by (a) a spherical
surface of radius Rs and (b) the Earth’s surface (Example 7-5).
(c) The power density Sav is related to the magnitude of the
e = E0 by
electric field |E|
Sav =
E02
,
2η0
where η0 = 377 (Ω) for air. Hence,
p
p
E0 = 2η0 Sav = 2 × 377 × 103 = 870
(V/m).
7-6.2 Plane Wave in a Lossy Medium
The expressions given by Eqs. (7.68) and (7.69) characterize
the electric and magnetic fields of an x-polarized plane wave
330
CHAPTER 7 PLANE-WAVE PROPAGATION
propagating along the z direction in a lossy medium with propagation constant γ = α + jβ . By extending these expressions
to the more general case of a wave with components along both
x and y, we have
Table 7-2 Power ratios in natural numbers and in decibels.
G
10x
4
2
1
0.5
0.25
0.1
10−3
e = x̂ Eex (z) + ŷ Eey (z) = (x̂ Ex0 + ŷ Ey0 )e−α z e− jβ z,
E(z)
(7.106a)
e = 1 (−x̂ Ey0 + ŷ Ex0 )e−α z e− jβ z ,
H(z)
ηc
(7.106b)
where ηc is the intrinsic impedance of the lossy medium.
Application of Eq. (7.100) gives
h
i
1
e×H
e∗
Re E×
2
ẑ(|Ex0 |2 + |Ey0 |2 ) −2α z
1
e
=
.
Re
2
ηc∗
Sav (z) =
the power ratio, particularly when numerical values of P1 /P2
are plotted against some variable of interest. If
(7.107)
G=
By expressing ηc in polar form as
ηc = |ηc |e jθη ,
(7.108)
Eq. (7.107) can be rewritten as
2
e
|E(0)|
e−2α z cos θη (W/m2 ),
Sav (z) = ẑ
2|ηc |
G [dB]
10x dB
6 dB
3 dB
0 dB
−3 dB
−6 dB
−10 dB
−30 dB
(7.109)
(lossy medium)
2 = [|E |2 + |E |2 ]1/2 is the magnitude of E(z)
e
e
where |E(0)|
x0
y0
at z = 0.
e and H(z)
e
◮ Whereas the fields E(z)
decay with z as e−α z ,
−2α z
.◭
the power density Sav decreases as e
When a wave propagates through a distance z = δs = 1/α ,
the magnitudes of its electric and magnetic fields decrease to
e−1 ≈ 37% of their initial values, and its average power density
decreases to e−2 ≈ 14% of its initial value.
7-6.3 Decibel Scale for Power Ratios
The unit for power P is watts (W). In many engineering
problems, the quantity of interest is the ratio of two power
levels, P1 and P2 , such as the incident and reflected powers
on a transmission line, and often the ratio P1 /P2 may vary
over several orders of magnitude. The decibel (dB) scale is
logarithmic, thereby providing a convenient representation of
P1
,
P2
(7.110)
then
G [dB] = 10 logG
P1
= 10 log
P2
(dB).
(7.111)
Table 7-2 provides a comparison between values of G and
the corresponding values of G [dB]. Even though decibels
are defined for power ratios, they can sometimes be used to
represent other quantities. For example, if P1 = V12 /R is the
power dissipated in a resistor R with voltage V1 across it at
time t1 , and P2 = V22 /R is the power dissipated in the same
resistor at time t2 , then
P1
G [dB] = 10 log
P2
2 V /R
= 10 log 12
V2 /R
V1
= 20 log
V2
= 20 log(g)
= g [dB],
(7.112)
where g = V1 /V2 is the voltage ratio. Note that for voltage (or
current) ratios the scale factor is 20 rather than 10, which
results in G [dB] = g [dB].
The attenuation rate, representing the rate of decrease of
the magnitude of Sav (z) as a function of propagation distance,
7-6 ELECTROMAGNETIC POWER DENSITY
331
e
Solution: From Example 7-4, |E(0)|
= |Ex0 | = 4.44 (mV/m),
α = 0.126 (Np/m), and ηc = 0.044 45◦ (Ω). Application of
Eq. (7.109) gives
is defined as
A = 10 log
Sav (z)
Sav (0)
= 10 log(e−2α z )
Sav (z) = ẑ
= −20α z log e = −8.68α z = −α [dB/m] z
(dB),
(7.113)
= ẑ
where
α [dB/m] = 8.68α
(Np/m).
(4.44 × 10−3)2 −0.252z
e
cos 45◦
2 × 0.044
= ẑ 0.16e−0.252z
(7.114)
Since Sav (z) is directly proportional to |E(z)|2 ,
|E(z)|2
|E(z)|
A = 10 log
(dB). (7.115)
= 20 log
|E(0)|2
|E(0)|
|E0 |2 −2α z
e
cos θη
2|ηc |
(mW/m2 ).
At z = 200 m, the incident power density is
Sav = ẑ (0.16 × 10−3e−0.252×200)
= 2.1 × 10−26
(W/m2 ).
Exercise 7-10: Convert the following values of the power
Example 7-6:
ratio G to decibels: (a) 2.3, (b) 4 × 103, and (c) 3 × 10−2.
Power Received by a
Submarine Antenna
Answer: (a) 3.6 dB, (b) 36 dB, (c) −15.2 dB. (See
A submarine at a depth of 200 m below the sea surface
uses a wire antenna to receive signal transmissions at 1 kHz.
Determine the power density incident upon the submarine
antenna due to the EM wave of Example 7-4.
EM
.)
Exercise 7-11: Find the voltage ratio g corresponding
to the following decibel values of the power ratio G:
(a) 23 dB, (b) −14 dB, and (c) −3.6 dB.
Answer: (a) 14.13, (b) 0.2, (c) 0.66. (See
EM
.)
Chapter 7 Summary
Concepts
• A spherical wave radiated by a source becomes approximately a uniform plane wave at large distances from
the source.
• The electric and magnetic fields of a transverse electromagnetic (TEM) wave are orthogonal to each other, and
both are perpendicular to the direction of wave travel.
• The magnitudes of the electric and magnetic fields of a
TEM wave are related by the intrinsic impedance of the
medium.
• Wave polarization describes the shape of the locus of
the tip of the E vector at a given point in space as a
function of time. The polarization state, which may be
linear, circular, or elliptical, is governed by the ratio of
the magnitudes of and the difference in phase between
the two orthogonal components of the electric field
vector.
• Media are classified as lossless, low-loss, quasiconducting, or good conducting on the basis of the ratio
ε ′′ /ε ′ = σ /ωε .
• Unlike the dc case, where the current flowing through
a wire is distributed uniformly across its cross section,
the ac case has most of the current is concentrated along
the outer perimeter of the wire.
• Power density carried by a plane EM wave traveling in
an unbounded medium is akin to the power carried by
the voltage/current wave on a transmission line.
332
CHAPTER 7 PLANE-WAVE PROPAGATION
Mathematical and Physical Models
Complex Permittivity
Maxwell’s Equations for Time-Harmonic Fields
εc = ε ′ − jε ′′
ε′ = ε
σ
ε ′′ =
ω
e=0
∇·E
e = − jω µ H
e
×E
∇×
e =0
∇·H
e
e = jωεc E
×H
∇×
Lossless Medium
√
k = ω µε
r
µ
η=
(Ω)
ε
ω
1
up = = √
k
µε
up
2π
λ=
=
k
f
(m/s)
(m)
Lossy Medium

1/2
s
′′ 2

 µε ′
ε

 1+
−
1
α =ω

 2
ε′

s
1/2
′′ 2

 µε ′
 1+ ε
+ 1
β =ω
′

 2
ε
ηc =
r
δs =
1
α
Wave Polarization
e
e = 1 k̂×
×E
H
η
e = −η k̂×
e
×H
E
Important Terms
µ
=
εc
r
µ
ε ′′ −1/2
1− j ′
ε′
ε
(Np/m)
(rad/m)
(Ω)
(m)
Power Density
Sav =
1
2
h
i
e×H
e∗
Re E×
(W/m2 )
Provide definitions or explain the meaning of the following terms:
attenuation constant α
attenuation rate A
auxiliary angle ψ0
average power density Sav
axial ratio
circular polarization
complex permittivity εc
dc and ac resistances
elliptical polarization
ellipticity angle χ
good conductor
guided wave
homogeneous wave equation
in phase
inclination angle
internal or surface impedance
intrinsic impedance η
LHC and RHC polarizations
linear polarization
lossy medium
low-loss dielectric
out of phase
phase constant β
phase velocity
polarization state
Poynting vector S
propagation constant γ
quasi-conductor
rotation angle γ
skin depth δs
spherical wave
surface resistance Rs
TEM wave
unbounded
unbounded wave
uniform plane wave
wave polarization
wavefront
wavenumber k
PROBLEMS
333
PROBLEMS
7.6 The electric field of a plane wave propagating in a
lossless, nonmagnetic, dielectric material with εr = 2.56 is
given by
Section 7-2: Propagation in Lossless Media
7.1 The magnetic field of a wave propagating through a
certain nonmagnetic material is given by
H = ẑ 30 cos(108t − 0.5y)
(mA/m).
E = ŷ 20 cos(6π × 109t − kz)
Determine:
(a) f , up , λ , k, and η .
Find:
∗(a) the direction of wave propagation,
(b) the phase velocity,
(b) the magnetic field H.
7.7 The magnetic field of a plane wave propagating in a
nonmagnetic material is given by
∗(c) the wavelength in the material,
H = x̂ 60 cos(2π × 107t + 0.1π y)
(d) the relative permittivity of the material, and
(e) the electric field phasor.
7.2 Write general expressions for the electric and magnetic
fields of a 1 GHz sinusoidal plane wave traveling in the
+y direction in a lossless nonmagnetic medium with relative
permittivity εr = 9. The electric field is polarized along the
x direction, its peak value is 6 V/m, and its intensity is 4 V/m
at t = 0 and y = 2 cm.
7.3 The electric field phasor of a uniform plane wave is given
e = ŷ 10e j0.2z (V/m). If the phase velocity of the wave is
by E
1.5 × 108 m/s and the relative permeability of the medium is
µr = 2.4, find:
∗(a) the wavelength,
(b) the frequency f of the wave,
(c) the relative permittivity of the medium, and
(d) the magnetic field H(z,t).
7.4 The electric field of a plane wave propagating in a
nonmagnetic material is given by
E = [ŷ 3 sin(π × 107t − 0.2π x)
+ ẑ4 cos(π × 107t − 0.2π x)]
(V/m)
Determine:
(a) the wavelength
(b) εr
(c) H
∗ 7.5 A wave radiated by a source in air is incident upon
a soil surface, where a part of the wave is transmitted into
the soil medium. If the wavelength of the wave is 60 cm in
air and 20 cm in the soil medium, what is the soil’s relative
permittivity? Assume the soil to be a very low-loss medium.
∗ Answer(s) available in Appendix E.
(V/m).
ẑ 30 cos(2π × 107t + 0.1π y)
(mA/m).
Determine:
∗ (a) the wavelength
(b) εr
(c) E
7.8 A 60 MHz plane wave traveling in the −x direction
in dry soil with relative permittivity εr = 4 has an electric
field polarized along the z direction. Assuming dry soil to be
approximately lossless, and given that the magnetic field has
a peak value of 10 (mA/m) and that its value was measured
to be 7 (mA/m) at t = 0 and x = −0.75 m, develop complete
expressions for the wave’s electric and magnetic fields.
Section 7-3: Wave Polarization
∗ 7.9 An RHC-polarized wave with a modulus of 2 (V/m) is
traveling in free space in the negative z direction. Write the
expression for the wave’s electric field vector given that the
wavelength is 6 cm.
7.10 For a wave characterized by the electric field
E(z,t) = x̂ ax cos(ω t − kz) + ŷay cos(ω t − kz + δ ),
identify the polarization state, determine the polarization
angles (γ , χ ), sketch the locus of E(0,t), and verify your
solution using Module 7.3 for each of the following cases.
(a) ax = 3 V/m, ay = 4 V/m, and δ = 0
(b) ax = 3 V/m, ay = 4 V/m, and δ = 180◦
(c) ax = 3 V/m, ay = 3 V/m, and δ = 45◦
(d) ax = 3 V/m, ay = 4 V/m, and δ = −135◦
334
CHAPTER 7 PLANE-WAVE PROPAGATION
7.11 The electric field of a uniform plane wave propagating
in free space is given by
e = (x̂ + jŷ)30e− jπ z/6
E
(V/m).
Specify the modulus and direction of the electric field intensity
at the z = 0 plane at t = 0, 5, and 10 ns.
∗ 7.12 The magnetic field of a uniform plane wave propagating
in a dielectric medium with εr = 36 is given by
e = 30(ŷ + jẑ)e− jπ x/6
H
(mA/m).
Specify the modulus and direction of the electric field intensity
at the x = 0 plane at t = 0 and 5 ns.
7.13 A linearly polarized plane wave of the form
e = x̂ ax e− jkz can be expressed as the sum of an RHCE
polarized wave with magnitude aR and an LHC-polarized
wave with magnitude aL . Prove this statement by finding
expressions for aR and aL in terms of ax .
∗ 7.14 The electric field of an elliptically polarized plane wave
is given by
E(z,t) = [−x̂ 10 sin(ω t − kz − 60◦)
+ ŷ30 cos(ω t − kz)]
(b) Animal tissue with µr = 1, εr = 12, and σ = 0.3 S/m at
100 MHz.
(c) Wood with µr = 1, εr = 3, and σ = 10−4 S/m at 1 kHz.
7.18 Dry soil is characterized by εr = 2.5, µr = 1, and
σ = 10−4 (S/m). At each of the following frequencies, determine if dry soil may be considered a good conductor, a quasiconductor, or a low-loss dielectric. Then calculate α , β , λ , µp ,
and ηc :
(a) 60 Hz
(b) 1 kHz
(c) 1 MHz
(d) 1 GHz
∗ 7.19 In a medium characterized by ε = 9, µ = 1, and
r
r
σ = 0.1 S/m, determine the phase angle by which the magnetic
field leads the electric field at 100 MHz.
7.20 Generate a plot for the skin depth δs versus frequency
for seawater for the range from 1 kHz to 10 GHz (use log-log
scales). The constitutive parameters of seawater are µr = 1,
εr = 80, and σ = 4 S/m.
∗ 7.21 Ignoring reflection at the air–soil boundary, if the am(V/m).
Determine:
(a) the polarization angles (γ , χ ), and
(b) the direction of rotation.
7.15 Compare the polarization states of each of the following
pairs of plane waves:
(a) Wave 1: E1 = x̂ 2 cos(ω t − kz) + ŷ2 sin(ω t − kz)
Wave 2: E2 = x̂ 2 cos(ω t + kz) + ŷ2 sin(ω t + kz)
(b) Wave 1: E1 = x̂ 2 cos(ω t − kz) − ŷ2 sin(ω t − kz)
Wave 2: E2 = x̂ 2 cos(ω t + kz) − ŷ2 sin(ω t + kz)
7.16 Plot the locus of E(0,t) for a plane wave with
E(z,t) = x̂ sin(ω t + kz) + ŷ2 cos(ω t + kz).
Determine the polarization state from your plot.
Sections 7-4: Propagation in a Lossy Medium
7.17 For each of the following combinations of parameters,
determine if the material is a low-loss dielectric, a quasiconductor, or a good conductor. Then calculate α , β , λ , up ,
and ηc .
∗ (a) Glass with µ = 1, ε = 5, and σ = 10−12 S/m at 10 GHz.
r
r
plitude of a 3 GHz incident wave is 10 V/m at the surface
of a wet soil medium, at what depth will it be down to
1 mV/m? Wet soil is characterized by µr = 1, εr = 9, and
σ = 5 × 10−4 S/m.
7.22 Ignoring reflection at the air–water boundary, if the
amplitude of a 1 GHz incident wave in air is 20 V/m at the
water surface, at what depth will it be down to 1 µ V/m? Water
has µr = 1 and at 1 GHz, εr = 80, and σ = 1 S/m.
∗ 7.23 The skin depth of a certain nonmagnetic conducting
material is 3 µ m at 5 GHz. Determine the phase velocity in
the material.
7.24 Based on wave attenuation and reflection measurements
conducted at 1 MHz, it was determined that the intrinsic
impedance of a certain medium is 28.1 45◦ (Ω) and the skin
depth is 2 m. Determine:
(a) the conductivity of the material,
(b) the wavelength in the medium, and
(c) the phase velocity.
∗ 7.25 The electric field of a plane wave propagating in a
nonmagnetic medium is given by
E = ẑ 25e−30x cos(2π × 109t − 40x)
Obtain the corresponding expression for H.
(V/m).
PROBLEMS
335
7.26 The magnetic field of a plane wave propagating in a
nonmagnetic medium is given by
H = ŷ 60e−10z cos(2π × 108t − 12z)
(mA/m).
7.34 A wave traveling in a nonmagnetic medium with εr = 9
is characterized by an electric field given by
E = [ŷ 3 cos(π × 107t + kx)
− ẑ2 cos(π × 107t + kx)]
Obtain the corresponding expression for E.
7.27 At 2 GHz, the conductivity of meat is on the order of
1 (S/m). When a material is placed inside a microwave oven
and the field is activated, the presence of the electromagnetic
fields in the conducting material causes energy dissipation in
the material in the form of heat.
(a) Develop an expression for the time-average power per
mm3 dissipated in a material of conductivity σ if the peak
electric field in the material is E0 .
(b) Evaluate the result for an electric field E0 = 4 × 104
(V/m).
Section 7-5: Current Flow in Conductors
H̃ = (x̂ − j4ẑ)e
e
(A/m).
Obtain time-domain expressions for the electric and magnetic
field vectors.
∗ 7.29 A rectangular copper block is 30 cm in height (along z).
In response to a wave incident upon the block from above,
a current is induced in the block in the positive x direction.
Determine the ratio of the ac resistance of the block to its dc
resistance at 1 kHz. The relevant properties of copper are given
in Appendix B.
7.30 Repeat Problem 7.29 at 10 MHz.
7.31 The inner and outer conductors of a coaxial cable have
radii of 0.5 cm and 1 cm, respectively. The conductors are
made of copper with εr = 1, µr = 1, and σ = 5.8 × 107 S/m.
The outer conductor is 0.5 mm thick. At 10 MHz:
(a) Are the conductors thick enough to be considered infinitely thick as far as the flow of current through them is
concerned?
(b) Determine the surface resistance Rs .
(c) Determine the ac resistance per unit length of the cable.
7.32 Repeat Problem 7.31 at 1 GHz.
Section 7-6: EM Power Density
∗ 7.33 The magnetic field of a plane wave traveling in air is
H = x̂ 50 sin(2π × 107t − ky)
Determine the direction of wave travel and average power
density carried by the wave.
7.35 The electric-field phasor of a uniform plane wave traveling downward in water is given by
e = x̂ 5e−0.2ze− j0.2z
E
(V/m),
where ẑ is the downward direction and z = 0 is the water
surface. If σ = 4 S/m,
(a) Obtain an expression for the average power density.
(b) Determine the attenuation rate.
∗ (c) Determine the depth at which the power density has been
7.28 In a nonmagnetic, lossy, dielectric medium, a 300 MHz
plane wave is characterized by the magnetic field phasor
−2y − j9y
(V/m).
(mA/m). Determine the
given by
average power density carried by the wave.
reduced by 40 dB.
7.36 The amplitudes of an elliptically polarized plane wave
traveling in a lossless, nonmagnetic medium with εr = 4 are
Hy0 = 3 (mA/m) and Hz0 = 4 (mA/m). Determine the average
power flowing through an aperture in the y–z plane if its area
is 20 m2 .
∗ 7.37 A wave traveling in a lossless, nonmagnetic medium
has an electric field amplitude of 24.56 V/m and an average
power density of 2.4 W/m2 . Determine the phase velocity of
the wave.
7.38 At microwave frequencies, the power density considered safe for human exposure is 1 (mW/cm2 ). A radar radiates
a wave with an electric field amplitude E that decays with
distance as E(R) = (3, 000/R) (V/m), where R is the distance
in meters. What is the radius of the unsafe region?
7.39 Consider the imaginary rectangular box shown in
Fig. P7.39.
(a) Determine the net power flux P(t) entering the box due to
a plane wave in air given by
E = x̂ E0 cos(ω t − ky)
(V/m).
∗ (b) Determine the net time-average power entering the box.
7.40 Repeat Problem 7.39 for a wave traveling in a lossy
medium in which
E = x̂ 100e−20y cos(2π × 109t − 40y)
H = −ẑ0.64e
−20y
(V/m),
cos(2π × 10 t − 40y − 36.85◦)
9
(A/m).
The box has dimensions a = 1 cm, b = 2 cm, and c = 0.5 cm.
336
CHAPTER 7 PLANE-WAVE PROPAGATION
∗ (a) Calculate the time-average electric energy density
z
(we )av =
b
a
c
1
T
Z T
0
we dt =
1
2T
Z T
ε E 2 dt.
0
(b) Calculate the time-average magnetic energy density
y
(wm )av =
1
T
Z T
0
wm dt =
1
2T
Z T
µ H 2 dt.
0
(c) Show that (we )av = (wm )av .
x
Figure P7.39 Imaginary rectangular box of Problems 7.39 and
7.40.
7.41 Given a wave with E = x̂ E0 cos(ω t − kz):
7.42 A team of scientists is designing a radar as a probe
for measuring the depth of the ice layer over the antarctic
land mass. In order to measure a detectable echo due to the
reflection by the ice-rock boundary, the thickness of the ice
sheet should not exceed three skin depths. If εr′ = 3 and
εr′′ = 10−2 for ice and if the maximum anticipated ice thickness
in the area under exploration is 1.2 km, what frequency range
is useable with the radar?
Chapter
8
Wave Reflection
and Transmission
Chapter Contents
8-1
TB15
8-2
8-3
8-4
TB16
8-5
8-6
8-7
8-8
8-9
8-10
8-11
Objectives
Upon learning the material presented in this chapter, you
should be able to:
EM Waves at Boundaries, 338
Wave Reflection and Transmission at Normal
Incidence, 338
Lasers, 347
Snell’s Laws, 349
Fiber Optics, 351
Wave Reflection and Transmission at Oblique
Incidence, 353
Bar-Code Readers, 358
Reflectivity and Transmissivity, 361
Waveguides, 364
General Relations for E and H, 366
TM Modes in Rectangular Waveguide, 367
TE Modes in Rectangular Waveguide, 370
Propagation Velocities, 371
Cavity Resonators, 374
Chapter 8 Summary, 377
Problems, 378
1. Characterize the reflection and transmission behavior of
plane waves incident upon plane boundaries for both
normal and oblique incidence.
2. Calculate the transmission properties of optical fibers.
3. Characterize wave propagation in a rectangular waveguide.
4. Determine the behavior of resonant modes inside a rectangular cavity.
337
EM Waves at Boundaries
Chapter 7. Finally, some of the power carried by the wave
traveling in water towards the submarine is intercepted by the
receiving antenna. The received power, Pr , is then delivered to
the receiver via a transmission line. The receiving properties
of antennas are covered in Chapter 9. In summary, each waverelated aspect of the transmission process depicted in Fig. 8-1
is treated in this book, starting with the transmitter and ending
with the receiver.
This chapter begins by examining the reflection and transmission properties of plane waves incident upon planar boundaries and concludes with sections on waveguides and cavity
resonators. Applications discussed along the way include fiber
and laser optics.
Figure 8-1 depicts the propagation path traveled by a signal
transmitted by a shipboard antenna and received by an antenna
on a submerged submarine. Starting from the transmitter
(denoted Tx in Fig. 8-1), the signal travels along a transmission
line to the transmitting antenna. The relationship between
the transmitter (generator) output power, Pt , and the power
supplied to the antenna is governed by the transmission-line
equations of Chapter 2. If the transmission line is approximately lossless and properly matched to the transmitting
antenna, then all of Pt is delivered to the antenna. If the antenna
itself is lossless too, it will convert all of the power Pt in the
guided wave provided by the transmission line into a spherical
wave radiated outward into space. The radiation process is the
subject of Chapter 9. From point 1, denoting the location of the
shipboard antenna, to point 2, denoting the point of incidence
of the wave onto the water’s surface, the signal’s behavior is
governed by the equations characterizing wave propagation in
lossless media, which was covered in Chapter 7. As the wave
impinges upon the air–water boundary, part of it is reflected
by the surface while another part gets transmitted across the
boundary into the water. The transmitted wave is refracted,
wherein its propagation direction moves closer toward the
vertical, compared with that of the direction of the incident
wave. Reflection and transmission processes are treated in
this chapter. Wave travel from point 3, representing a point
just below the water surface, to point 4, which denotes the
location of the submarine antenna, is subject to the laws of
wave propagation in lossy media, which also was treated in
8-1 Wave Reflection and Transmission at
Normal Incidence
We know from Chapter 2 that, when a guided wave encounters a junction between two transmission lines with different
characteristic impedances, the incident wave is partly reflected
back toward the source and partly transmitted across the
junction onto the other line. The same happens to a uniform
plane wave when it encounters a boundary between two
material half-spaces with different characteristic impedances.
In fact, the situation depicted in Fig. 8-2(b) has an exact
analogue in the transmission-line configuration of Fig. 8-2(a).
The boundary conditions governing the relationships between
the electric and magnetic fields in Fig. 8-2(b) map one-to-one
onto those we developed in Chapter 2 for the voltages and
currents on the transmission line.
For convenience, we divide our treatment of wave reflection
by and transmission through planar boundaries into two parts.
In this section, we confine our discussion to the normalincidence case depicted in Fig. 8-3(a), and in Sections 8-2
to 8-4, we examine the more general oblique-incidence case
depicted in Fig. 8-3(b). We will show the basis for the analogy
between the transmission-line and plane-wave configurations
so that we may use transmission-line equivalent models, tools
(e.g., Smith chart), and techniques (e.g., quarter-wavelength
matching) to expeditiously solve plane-wave problems.
Before proceeding, however, we should explain the notion
of rays and wavefronts and the relationship between them,
as both are used throughout this chapter to represent electromagnetic waves. A ray is a line representing the direction of
flow of electromagnetic energy carried by a wave; therefore,
it is parallel to the propagation unit vector k̂. A wavefront is
a surface across which the phase of a wave is constant; it is
perpendicular to the wavevector k̂. Hence, rays are perpendicular to wavefronts. The ray representation of wave incidence,
reflection, and transmission shown in Fig. 8-3(b) is equivalent
Transmitter
antenna
1
Pt
Tx
2
Air
3
Water
Receiver
antenna
4
Pr
Rx
Figure 8-1 Signal path between a shipboard transmitter (Tx)
and a submarine receiver (Rx).
338
8-1 WAVE REFLECTION AND TRANSMISSION AT NORMAL INCIDENCE
339
Incident plane wave
Transmission line 1
Transmission line 2
Incident wave
Z01
Reflected wave
Transmitted
wave
Transmitted plane wave
Reflected plane wave
Z02
Medium 1
η1
z=0
(a) Boundary between transmission lines
Medium 2
η2
z=0
(b) Boundary between different media
Figure 8-2 Discontinuity between two different transmission lines is analogous to that between two dissimilar media.
Incident wave
Reflected wave
Reflected
wave
Transmitted wave
Transmitted
wave
θr
θr
θt
θt
θi
Medium 1
η1
Medium 2
η2
Incident
wave
Medium 1
η1
(a) Normal incidence
θi
Medium 2
η2
(b) Ray representation of
oblique incidence
Medium 1
η1
Medium 2
η2
(c) Wavefront representation of
oblique incidence
Figure 8-3 Ray representation of wave reflection and transmission at (a) normal incidence and (b) oblique incidence. (c) Wavefront
representation of oblique incidence.
to the wavefront representation depicted in Fig. 8-3(c). The
two representations are complementary; the ray representation
is easier to use in graphical illustrations, whereas the wavefront
representation provides greater physical insight into what happens to a wave when it encounters a discontinuous boundary.
8-1.1 Boundary between Lossless Media
A planar boundary located at z = 0 (Fig. 8-4(a)) separates
two lossless, homogeneous, dielectric media. Medium 1 has
permittivity ε1 and permeability µ1 and fills the half-space
z ≤ 0. Medium 2 has permittivity ε2 and permeability µ2 and
fills the half-space z ≥ 0. An x-polarized plane wave, with
electric and magnetic fields (Ei , Hi ) propagates in medium 1
along direction k̂i = ẑ toward medium 2. Reflection and transmission at the boundary at z = 0 result in a reflected wave, with
electric and magnetic fields (Er , Hr ), traveling along direction
k̂r = −ẑ in medium 1, and a transmitted wave, with electric
and magnetic fields (Et , Ht ), traveling along direction k̂t = ẑ
in medium 2. On the basis of the formulations developed in
Sections 7-2 and 7-3 for plane waves, the three waves are
described in phasor form by:
340
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Ei
x
Et
kˆ i
Hi
kˆ t
Ht
Er
z
y
kˆ r
p
√
η1 = pµ1 /ε1 , and those for medium 2 are k2 = ω µ2 ε2 and
η2 = µ2 /ε2 .
The amplitude E0i is imposed by the source responsible for
generating the incident wave; therefore, it is assumed known.
Our goal is to relate E0r and E0t to E0i . We do so by applying
boundary conditions for the total electric and magnetic fields
at z = 0. According to Table 6-2, the tangential component of
the total electric field is always continuous across a boundary
between two contiguous media, and in the absence of current
sources at the boundary, the same is true for the total magnetic
field. In the present case, the electric and magnetic fields of the
incident, reflected, and transmitted waves are all tangential to
the boundary.
e 1 (z) in medium 1 is the sum of the
The total electric field E
electric fields of the incident and reflected waves, and a similar
e 1 (z). Hence,
statement applies to the magnetic field H
Hr
Medium 2 (ε2, μ2)
z=0
(a) Boundary between dielectric media
Medium 1 (ε1, μ1)
Z01
Z02
Infinite line
z=0
(b) Transmission-line analogue
Figure 8-4
The two dielectric media separated by the
x–y plane in (a) can be represented by the transmission-line
analogue in (b).
Medium 1
e 1 (z) = E
e i (z) + E
e r (z) = x̂(E0i e− jk1 z + E0r e jk1 z ),
E
e 1 (z) = H
e i (z) + H
e r (z) = ŷ 1 (E0i e− jk1 z − E0r e jk1 z ). (8.4b)
H
η1
With only the transmitted wave present in medium 2, the total
fields are
Medium 2
e 2 (z) = E
e t (z) = x̂E0t e− jk2 z ,
E
t
e 2 (z) = H
e t (z) = ŷ E0 e− jk2 z .
H
η2
Incident Wave
e i (z) = x̂E0i e− jk1 z ,
E
ei
(8.1a)
i
E
E (z)
e i (z) = ẑ×
×
= ŷ 0 e− jk1 z .
H
η1
η1
Reflected Wave
e r (z) = x̂E0r e jk1 z ,
E
e r (z)
Er
E
e r (z) = (−ẑ)×
×
= −ŷ 0 e jk1 z .
H
η1
η1
(8.2b)
Transmitted Wave
e t (z) = x̂E0t e− jk2 z ,
E
e t (z)
Et
E
e t (z) = ẑ×
×
= ŷ 0 e− jk2 z .
H
η2
η2
(8.3a)
(8.5a)
(8.5b)
At the boundary (z = 0), the tangential components of the
electric and magnetic fields are continuous. Hence,
(8.1b)
(8.2a)
(8.4a)
e 1 (0) = E
e 2 (0) or E0i + E0r = E0t ,
E
e 1 (0) = H
e 2 (0) or
H
Et
E0i E0r
−
= 0.
η1 η1
η2
(8.6a)
(8.6b)
Solving these equations for E0r and E0t in terms of E0i gives
η2 − η1
r
E0 =
E0i = ΓE0i ,
(8.7a)
η2 + η1
2η2
E0i = τ E0i ,
(8.7b)
E0t =
η2 + η1
where
(8.3b)
The quantities E0i , E0r , and E0t are, respectively, the amplitudes
of the incident, reflected, and transmitted electric fields at
z = 0 (the boundary between the two media). The wavenumber
√
and intrinsic impedance of medium 1 are k1 = ω µ1 ε1 and
Γ=
E0r
η2 − η1
=
η2 + η1
E0i
(normal incidence),
(8.8a)
τ=
E0t
2η2
=
i
η2 + η1
E0
(normal incidence).
(8.8b)
8-1 WAVE REFLECTION AND TRANSMISSION AT NORMAL INCIDENCE
The quantities Γ and τ are called the reflection and transmission coefficients. For lossless dielectric media, η1 and η2 are
real; consequently, both Γ and τ are real also. As we see in
Section 8-1.4, the expressions given by Eqs. (8.8a) and (8.8b)
are equally applicable when the media are conductive, even
though in that case η1 and η2 may be complex. Hence, Γ and τ
may be complex as well. From Eqs. (8.8a) and (8.8b), it is
easily shown that Γ and τ are interrelated as
τ = 1+Γ
(normal incidence).
(8.9)
For nonmagnetic media,
η0
η1 = √ ,
εr1
η0
and η2 = √ ,
εr2
where η0 is the intrinsic impedance of free space. In this case,
Eq. (8.8a) may be expressed as
√
√
εr − εr
Γ= √ 1 √ 2
εr1 + εr2
(nonmagnetic media).
(8.10)
8-1.2 Transmission-Line Analogue
The transmission-line configuration shown in Fig. 8-4(b)
consists of a lossless transmission line with characteristic
impedance Z01 connected at z = 0 to an infinitely long lossless
transmission line with characteristic impedance Z02 . The input
impedance of an infinitely long line is equal to its characteristic
impedance. Hence, at z = 0, the voltage reflection coefficient
(looking toward the boundary from the vantage point of the
first line) is
Z02 − Z01
,
Γ=
Z02 + Z01
which is identical in form to Eq. (8.8a). The analogy between
plane waves and waves on transmission lines does not end
here. To demonstrate the analogy further, equations pertinent
to the two cases are summarized in Table 8-1. Comparison of
the two columns shows that there is a one-to-one correspone β , Z0 ) and
dence between the transmission-line quantities (Ve , I,
e H,
e k, η ).
the plane-wave quantities (E,
◮ This correspondence allows us to use the techniques
developed in Chapter 2, including the Smith-chart method
for calculating impedance transformations, to solve planewave propagation problems. ◭
The simultaneous presence of incident and reflected waves
in medium 1 (Fig. 8-4(a)) gives rise to a standing-wave pattern.
341
By analogy with the transmission-line case, the standing-wave
ratio in medium 1 is defined as
S=
1 + |Γ|
|Ee1 |max
=
.
e
1
− |Γ|
|E1 |min
(8.15)
◮ If the two media have equal impedances (η1 = η2 ),
then Γ = 0 and S = 1, and if medium 2 is a perfect
conductor with η2 = 0 (which is equivalent to a shortcircuited transmission line), then Γ = −1 and S = ∞. ◭
The distance from the boundary to where the magnitude of the
electric field intensity in medium 1 is a maximum (denoted
lmax ) is described by the same expression as that given by
Eq. (2.70) for the voltage maxima on a transmission line,
namely
−z = lmax =
θr + 2nπ
θr λ1 nλ1
+
=
,
2k1
4π
2
(8.16)
n = 1, 2, . . . if θr < 0,
n = 0, 1, 2, . . . if θr ≥ 0,
where λ1 = 2π /k1 and θr is the phase angle of Γ (i.e.,
Γ = |Γ|e jθr , and θr is bounded in the range −π < θr ≤ π ).
The expression for lmax is valid not only when the two media
are lossless dielectrics, but also when medium 1 is a low-loss
dielectric. Moreover, medium 2 may be either a dielectric or a
conductor. When both media are lossless dielectrics, θr = 0 if
η2 > η1 , and θr = π if η2 < η1 .
The spacing between adjacent maxima is λ1 /2, and the
spacing between a maximum and the nearest minimum is
λ1 /4. The electric-field minima occur at
lmin =
lmax + λ1/4, if lmax < λ1 /4,
lmax − λ1/4, if lmax ≥ λ1 /4.
(8.17)
8-1.3 Power Flow in Lossless Media
Medium 1 in Fig. 8-4(a) is host to the incident and reflected
waves, which together comprise the total electric and magnetic
e 1 (z) and H
e 1 (z) given by Eqs. (8.11a) and (8.12a) of
fields E
Table 8-1. Using Eq. (7.100), the net average power density
342
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Table 8-1 Analogy between plane-wave equations for normal incidence and transmission-line equations with both under lossless
conditions.
Plane Wave (Fig. 8-4(a))
e 1 (z) = x̂E i (e− jk1 z + Γe jk1 z )
E
0
e 1 (z) = ŷ
H
E0i
η1
(e− jk1 z − Γe jk1 z )
e 2 (z) = x̂τ E i e− jk2 z
E
0
(8.11a)
(8.12a)
(8.13a)
E0i
e− jk2 z
η2
Γ = (η2 − η1 )/(η2 + η1 )
e 2 (z) = ŷτ
H
Transmission Line (Fig. 8-4(b))
(8.14a)
Ve1 (z) = V0+ (e− jβ1 z + Γe jβ1 z )
(8.11b)
Ve2 (z) = τV0+ e− jβ2 z
(8.13b)
Ie1 (z) =
k1 = ω µ1 ε1 ,
p
η1 = µ1 /ε1 ,
(e− jβ1 z − Γe jβ1 z )
(8.12b)
V0+
e− jβ2 z
Z02
Γ = (Z02 − Z01 )/(Z02 + Z01 )
(8.14b)
τ = 1+Γ
√
√
k2 = ω µ2 ε2
p
η2 = µ2 /ε2
β1 = ω µ1 ε1 ,
√
β2 = ω µ2 ε2
Z01 and Z02 depend on
transmission-line parameters
flowing in medium 1 is
medium 2 is
e 1 (z)×
e ∗1 (z)]
×H
Sav1 (z) = 21 Re[E
E i∗
= 21 Re x̂E0i (e− jk1 z + Γe jk1 z )ŷ 0 (e jk1 z − Γ∗ e− jk1 z )
η1
= ẑ
Z01
Ie2 (z) = τ
τ = 1+Γ
√
V0+
|E0i |2
(1 − |Γ|2),
2η1
(8.18)
which is analogous to Eq. (2.106) for the lossless transmissionline case. The first and second terms inside the bracket in
Eq. (8.18) represent the average power density of the incident
and reflected waves, respectively. Thus,
Sav1 = Siav + Srav
(8.19a)
∗
e 2 (z)×
e 2 (z)]
×H
Sav2 (z) = 21 Re[E
i∗
|E i |2
∗ E0
i − jk2 z
jk2 z
1
= 2 Re x̂τ E0 e
= ẑ|τ |2 0 .
× ŷτ
e
η2
2η2
(8.20)
Through the use of Eqs. (8.8a and b), it can be easily shown
that for lossless media
1 − Γ2
τ2
=
,
η2
η1
(lossless media)
(8.21)
which leads to
Sav1 = Sav2 .
with
Siav
|E i |2
= ẑ 0 ,
2η1
Srav = −ẑ|Γ|2
|E0i |2
= −|Γ|2 Siav .
2η1
(8.19b)
(8.19c)
Even though Γ is purely real when both media are lossless
dielectrics, we chose to treat it as complex, thereby providing
in Eq. (8.19c) an expression that is also valid when medium 2
is conducting.
The average power density of the transmitted wave in
This result is expected from considerations of power conservation.
Example 8-1: Radar Radome Design
A 10 GHz aircraft radar uses a narrow-beam scanning antenna
mounted on a gimbal behind a dielectric radome, as shown in
Fig. 8-5. Even though the radome shape is far from planar, it is
approximately planar over the narrow extent of the radar beam.
If the radome material is a lossless dielectric with εr = 9 and
8-1 WAVE REFLECTION AND TRANSMISSION AT NORMAL INCIDENCE
Antenna beam
Radar
Antenna
Dielectric
radome
d
Figure 8-5
Antenna beam “looking” through an aircraft
radome of thickness d (Example 8-1).
Incident wave
Radome
Transmitted wave
343
on an expanded scale. The incident wave can be approximated
as a plane wave propagating in medium 1 (air) with intrinsic
impedance η0 . Medium 2 (the radome) is of thickness d and
intrinsic impedance ηr , and medium 3 (air) is semi-infinite
with intrinsic impedance η0 . Figure 8-6(b) shows an equivalent transmission-line model with z = 0 selected to coincide
with the outside surface of the radome, and the load impedance
ZL = η0 represents the input impedance of the semi-infinite air
medium to the right of the radome.
For the radome to “appear” transparent to the incident wave,
the reflection coefficient must be zero at z = −d, thereby
guaranteeing total transmission of the incident power into
medium 3. Since ZL = η0 in Fig. 8-6(b), no reflection takes
place at z = −d if Zin = η0 , which can be realized by choosing
d = nλ2/2 [see Section 2-8.4], where λ2 is the wavelength
in medium 2 and n is a positive integer. At 10 GHz, the
wavelength in air is λ0 = c/ f = 3 cm, while in the radome
material it is
λ0
3 cm
λ2 = √ =
= 1 cm.
εr
3
Hence, by choosing d = 5λ2 /2 = 2.5 cm, the radome is both
nonreflecting and structurally stable.
Medium 1 (air)
η0
Medium 2
ηr
z = −d
Medium 3 (air)
η0
Line 2
Line 1
Zin
Yellow Light Incident
upon a Glass Surface
z=0
(a)
Z01 = η0
Example 8-2:
Z02 = ηr
z = −d
ZL = η0
z=0
(b)
Figure 8-6 (a) Planar section of the radome of Fig. 8-5 at an
expanded scale and (b) its transmission-line equivalent model
(Example 8-1).
µr = 1, choose its thickness d such that the radome appears
transparent to the radar beam. Structural integrity requires d to
be greater than 2.3 cm.
Solution: Figure 8-6(a) shows a small section of the radome
A beam of yellow light with wavelength 0.6 µ m is normally
incident in air upon a glass surface. Assume the glass is
sufficiently thick as to ignore its back surface. If the surface
is situated in the plane z = 0 and the relative permittivity of
glass is 2.25, determine:
(a) the locations of the electric field maxima in medium 1
(air),
(b) the standing-wave ratio, and
(c) the fraction of the incident power transmitted into the
glass medium.
Solution: (a) We begin by determining the values of η1 , η2 ,
and Γ:
r
r
µ1
µ0
η1 =
=
≈ 120π (Ω),
ε1
ε0
r
r
µ2
µ0 1
120π
η2 =
=
·√ ≈√
= 80π (Ω),
ε2
ε0
εr
2.25
η2 − η1 80π − 120π
Γ=
=
= −0.2.
η2 + η1 80π + 120π
344
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Hence, |Γ| = 0.2 and θr = π . From Eq. (8.16), the electric-field
magnitude is maximum at
lmax =
θr λ1
λ1 λ1
λ1
+n =
+n
4π
2
4
2
Ei
x
(n = 0, 1, 2, . . .)
kˆ i
Hi
with λ1 = 0.6 µ m.
E
kˆ r
(c) The fraction of the incident power transmitted into the
glass medium is equal to the ratio of the transmitted power
density given by Eq. (8.20) to the incident power density,
i
Sav
= |E0i |2 /2η1 :
,
i 2
|E0i |2
Sav2
η1
2 |E0 |
= τ2
.
=τ
i
Sav
2η2
2η1
η2
In view of Eq. (8.21),
In Section 8-1.1, we considered a plane wave in a lossless
medium incident normally on a planar boundary of another
lossless medium. We now generalize our expressions to lossy
media. In a medium with constitutive parameters (ε , µ , σ ),
the propagation constant γ = α + jβ and the intrinsic impedance ηc are both complex. General expressions for α , β ,
and ηc are given by Eqs. (7.66a), (7.66b), and (7.70), respectively, and approximate expressions are given in Table 7-2
for the special cases of low-loss media and good conductors.
If media 1 and 2 have constitutive parameters (ε1 , µ1 , σ1 )
and (ε2 , µ2 , σ2 ) (Fig. 8-7), then expressions for the electric
and magnetic fields in media 1 and 2 can be obtained from
Eqs. (8.11) through (8.14) of Table 8-1 by replacing jk with γ
and η with ηc . Thus,
Medium 1
Medium 2
i −γ2 z
e 2 (z) = x̂τ E0 e
E
,
i
e 2 (z) = ŷτ E0 e−γ2 z .
H
ηc2
ηc1
ηc2
Medium 1 (ε1, μ1, σ1)
Medium 2 (ε2, μ2, σ2)
z=0
(a) Boundary between dielectric media
Z01 = ηc1
Z02 = ηc2
z=0
(b) Transmission-line analogue
8-1.4 Boundary between Lossy Media
i
e 1 (z) = ŷ E0 (e−γ1 z − Γeγ1 z ),
H
ηc1
Hr
Infinite line
Sav2
= 1 − |Γ|2 = 1 − (0.2)2 = 0.96, or 96%.
i
Sav
e 1 (z) = x̂E0i (e−γ1 z + Γeγ1 z ),
E
z
y
1 + |Γ| 1 + 0.2
=
= 1.5.
1 − |Γ| 1 − 0.2
kˆ t
Ht
r
(b)
S=
Et
Figure 8-7 Normal incidence at a planar boundary between
two lossy media.
Here, γ1 = α1 + jβ1 , γ2 = α2 + jβ2 , and
Γ=
ηc2 − ηc1
,
ηc2 + ηc1
τ = 1+Γ =
2ηc2
.
ηc2 + ηc1
(8.24a)
(8.24b)
Because ηc1 and ηc2 are, in general, complex, Γ and τ may be
complex as well.
(8.22a)
Example 8-3:
(8.22b)
(8.23a)
(8.23b)
Normal Incidence on a Metal
Surface
A 1 GHz x-polarized plane wave traveling in the +z direction
is incident from air upon a copper surface. The air-to-copper
interface is at z = 0, and copper has εr = 1, µr = 1, and
σ = 5.8 × 107 S/m. If the amplitude of the electric field of
the incident wave is 12 (mV/m), obtain expressions for the
8-1 WAVE REFLECTION AND TRANSMISSION AT NORMAL INCIDENCE
345
instantaneous electric and magnetic fields in the air medium.
Assume the metal surface to be several skin depths deep.
ωt = 3π/2
Solution: In medium 1 (air), α = 0,
–z
ωt = 5π/4
ωt = 0
–λ
–3λ
4
24 (mV/m)
0
4
ωt = π/4
Conductor
2π × 109 20π
ω
=
=
(rad/m),
c
3 × 108
3
2π
= 0.3 m.
η1 = η0 = 377 (Ω),
λ=
k1
β = k1 =
E1(z, t)
At f = 1 GHz, copper is an excellent conductor because
5.8 × 107
ε ′′
σ
=
= 1 × 109 ≫ 1.
=
′
ε
ωεr ε0
2π × 109 × (10−9/36π )
–24 (mV/m)
ωt = π/2
Use of Eq. (7.77c) gives
r
ηc2 = (1 + j)
ωt = π
64 (μA/m)
(mΩ).
Since ηc2 is so small compared to η0 = 377 (Ω) for air, the
copper surface acts, in effect, like a short circuit. Hence,
Γ=
–z
–λ
–λ
2
ηc2 − η0
≈ −1.
ηc2 + η0
0
Conductor
1/2
π fµ
π × 109 × 4π × 10−7
= (1 + j)
σ
5.8 × 107
= 8.25(1 + j)
H1(z, t)
–64 (μA/m)
ωt = 0
Upon setting Γ = −1 in Eqs. (8.11) and (8.12) of Table 8-1,
we obtain
Figure 8-8 Wave patterns for fields E1 (z,t) and H1 (z,t) of
Example 8-3.
e 1 (z) = x̂E0i (e− jk1 z − e jk1 z ) = −x̂ j2E0i sin k1 z,
E
i
i
e 1 (z) = ŷ E0 (e− jk1 z + e jk1 z ) = ŷ2 E0 cos k1 z.
H
η1
η1
(8.25a)
(8.25b)
With E0i = 12 (mV/m), the instantaneous fields associated with
these phasors are
e 1 (z) e jω t ]
E1 (z,t) = Re[E
= x̂ 2E0i sin k1 z sin ω t
= x̂ 24 sin(20π z/3) sin(2π × 109t) (mV/m),
e 1 (z) e jω t ]
H1 (z,t) = Re[H
= ŷ 2
E0i
cosk1 z cos ω t
η1
= ŷ 64 cos(20π z/3) cos(2π × 109t) (µ A/m).
Plots of the magnitude of E1 (z,t) and H1 (z,t) are shown in
Fig. 8-8 as a function of negative z for various values of ω t.
The wave patterns exhibit a repetition period of λ /2, and E
and H are in phase quadrature (90◦ phase shift) in both space
and time. This behavior is identical with that for voltage and
current waves on a shorted transmission line.
What boundary conditions
were used in the derivations of the expressions for Γ
and τ ?
Concept Question 8-1:
In the radar radome design of
Example 8-1, all the incident energy in medium 1 ends up
getting transmitted into medium 3, and vice versa. Does
this imply that no reflections take place within medium 2?
Explain.
Concept Question 8-2:
346
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Module 8.1 Incidence on Perfect Conductor Observe the standing-wave pattern created by the combination of a wave
incident normally onto the plane surface of a conductor and its reflection.
Explain on the basis of boundary conditions why it is necessary that Γ = −1 at the
boundary between a dielectric and a perfect conductor.
Concept Question 8-3:
Exercise 8-1: To eliminate reflections of normally inci-
dent plane waves, a dielectric slab of thickness d and
relative permittivity εr2 is to be inserted between two
semi-infinite media with relative permittivities εr1 = 1 and
εr3 = 16. Use the quarter-wave transformer technique to
select d and εr2 . Assume f = 3 GHz.
εr2 = 4 and d = (1.25 + 2.5n) (cm), with
n = 0, 1, 2, . . . . (See EM .)
Answer:
Exercise 8-2: Express the normal-incidence reflection
coefficient at the boundary between two nonmagnetic,
conducting media in terms of their complex permittivities.
Answer: For incidence in medium 1 (ε1 , µ0 , σ1 ) onto
medium 2 (ε2 , µ0 , σ2 ),
√
√
εc − εc
Γ= √ 1 √ 2 ,
εc1 + εc2
with εc1 = (ε1 − jσ1 /ω ) and εc2 = (ε2 − jσ2 /ω ). (See
EM
.)
Exercise 8-3: Obtain expressions for the average power
densities in media 1 and 2 for the fields described by
Eqs. (8.22a) through (8.23b), assuming medium 1 is
slightly lossy with ηc1 approximately real.
Answer: (See
EM
.)
|E0i |2 −2α1 z
− |Γ|2 e2α1 z ,
e
2ηc1
i 2
1
−2α2 z
2 |E0 |
e
= ẑ|τ |
.
Re
2
ηc∗2
Sav1 = ẑ
Sav2
8-1 WAVE REFLECTION AND TRANSMISSION AT NORMAL INCIDENCE
Technology Brief 15: Lasers
Lasers are used in CD and DVD players, bar-code
readers, eye surgery, and multitudes of other systems
and applications (Fig. TF15-1).
◮ A laser—acronym for Light Amplification by
Stimulated Emission of Radiation—is a source
of monochromatic (single wavelength), coherent
(uniform wavefront), narrow-beam light. ◭
This is in contrast with other sources of light (such
as the sun or a light bulb), which usually encompass
waves of many different wavelengths with random phase
(incoherent). A laser source generating microwaves is
called a maser. The first maser was built in 1953 by
Charles Townes, and the first laser was constructed in
1960 by Theodore Maiman.
Basic Principles
Despite its complex quantum-mechanical structure, an
atom can be conveniently modeled as a nucleus (containing protons and neutrons) surrounded by a cloud
of electrons. Associated with the atom or molecule
of any given material is a specific set of quantized
(discrete) energy states (orbits) that the electrons can
occupy. Supply of energy (in the form of heat, exposure to intense light, or other means) by an external
source can cause an electron to move from a lower
energy state to a higher energy (excited) state. Exciting the atoms is called pumping because it leads to
347
increasing the population of electrons in higher states
(Fig. TF15-2(a)). Spontaneous emission of a photon
(light energy) occurs when the electron in the excited
state moves to a lower state (Fig. TF15-2(b)), and
stimulated emission (Fig. TF15-2(c)) happens when an
emitted photon “entices” an electron in an excited state of
another atom to move to a lower state, thereby emitting
a second photon of identical energy, wavelength, and
wavefront (phase).
Principle of Operation
◮ Highly amplified stimulated emission is called
lasing. ◭
The lasing medium can be solid, liquid, or gas. Laser
operation is illustrated in Fig. TF15-3 for a ruby crystal
surrounded by a flash tube (similar to a camera flash).
A perfectly reflecting mirror is placed on one end of
the crystal and a partially reflecting mirror on the other
end. Light from the flash tube excites the atoms; some
undergo spontaneous emission, generating photons that
cause others to undergo stimulated emission; photons
moving along the axis of the crystal bounce back and
forth between the mirrors, causing additional stimulated
emission (i.e., amplification), with only a fraction of the
photons exiting through the partially reflecting mirror.
◮ Because all of the stimulated photons are identical, the light wave generated by the laser is of a
single wavelength. ◭
Figure TF15-1 A few examples of laser applications.
348
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Incident Electron
energy or
photon
Orbit of
ground state
Photon
Original photon
Photon
Stimulated photon
Nucleus
Orbit of
excited state
(a) Pumping electron to excited state
(b) Spontaneous emission
(c) Stimulated emission
Figure TF15-2 Electron excitation and photon emission.
Excitation energy
(e.g., flash tube)
Perfectly
reflecting mirror
Partially
reflecting mirror
Laser light
Amplifying medium
Figure TF15-3 Laser schematic.
Wavelength (Color) of Emitted Light
The atom of any given material has unique energy
states. The difference in energy between the excited
high energy state and the stable lower energy state
determines the wavelength of the emitted photons
(EM wave). Through proper choice of lasing material,
monochromatic waves can be generated with wavelengths in the ultraviolet, visible, infrared or microwave
bands.
8-2 SNELL’S LAWS
349
8-2 Snell’s Laws
In the preceding sections, we examined reflection and transmission of plane waves that are normally incident upon a
planar interface between two different media. We now consider the oblique-incidence case depicted in Fig. 8-9, and for
simplicity, we assume all media to be lossless. The z = 0 plane
forms the boundary between media 1 and 2 with constitutive
parameters (ε1 , µ1 ) and (ε2 , µ2 ), respectively. The two lines
in Fig. 8-9 with direction k̂i represent rays drawn normal to
the wavefront of the incident wave, and those along directions
k̂r and k̂t are similarly associated with the reflected and
transmitted waves. The angles of incidence, reflection, and
transmission (or refraction), which are defined with respect
to the normal to the boundary (the z axis), are θi , θr , and θt ,
respectively. These three angles are interrelated by Snell’s
laws, which we derive shortly by considering the propagation
of the wavefronts of the three waves. Rays of the incident
wave intersect the boundary at O and O′ . Here Ai O represents
a constant-phase wavefront of the incident wave. Likewise,
Ar O′ and At O′ are constant-phase wavefronts of the reflected
and transmitted waves, respectively (Fig. 8-9). The incident
and reflected waves propagate
in medium 1 with the same
√
wave in
phase velocity up1 = 1/ µ1 ε1 , while the transmitted
√
medium 2 propagates with a velocity up2 = 1/ µ2 ε2 . The time
it takes for the incident wave to travel from Ai to O′ is the same
as the time it takes for the reflected wave to travel from O to Ar ,
and also the time it takes the transmitted wave to travel from O
to At . Since time equals distance divided by velocity, it follows
that
A i O′
OAr
OAt
=
=
.
(8.26)
up1
up1
up2
From the geometries of the three right triangles in Fig. 8-9, we
deduce that
Ai O′ = OO′ sin θi ,
(8.27a)
OAr = OO′ sin θr ,
(8.27b)
OAt = OO′ sin θt .
(8.27c)
Use of these expressions in Eq. (8.26) leads to
θi = θr ,
up
sin θt
= 2
sin θi
up1
(Snell’s law of reflection)
r
µ1 ε1
=
.
µ2 ε2
(8.28a)
(8.28b)
(Snell’s law of refraction)
◮ Snell’s law of reflection states that the angle of reflection equals the angle of incidence, and Snell’s law of
refraction provides a relation between sin θt and sin θi in
terms of the ratio of the phase velocities. ◭
The index of refraction of a medium, n, is defined as the
ratio of the phase velocity in free space (i.e., the speed of
light c) to the phase velocity in the medium. Thus,
x
Reflected wave ˆ
kr
ˆkr
ˆkt
Transmitted
wave
ˆkt
O'
Ar
Ai
θr
θi
At θt
O
z
ˆki
n=
r
√
µε
= µr εr .
µ0 ε0
(8.29)
(index of refraction)
In view of Eq. (8.29), Eq. (8.28b) may be rewritten as
r
µr1 εr1
sin θt
n1
=
=
.
(8.30)
sin θi
n2
µr2 εr2
For nonmagnetic materials, µr1 = µr2 = 1, in which case
ˆki
Incident wave
Medium 1 (ε1, μ1)
Medium 2 (ε2, μ2)
Figure 8-9 Wave reflection and refraction at a planar boundary
between different media.
c
=
up
sin θt
n1
=
=
sin θi
n2
r
εr1
η2
=
εr2
η1
(for µ1 = µ2 ).
(8.31)
Usually, materials with higher densities have higher permittivities. Air, with µr = εr = 1, has an index
√ of refraction n0 = 1.
Since for nonmagnetic materials n = εr , a material is often
350
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Inward refraction
Outward refraction
θt > θi
θr
θi
n1
θr
θi
θt
n2
n1
(a) n1 < n2
the critical angle θc and is obtained from Eq. (8.30) as
n2
sin θt
n1
r
εr2
=
εr1
sin θc =
θt
θt =π /2
=
n2
n1
(for µ1 = µ2 ).
(8.32a)
(8.32b)
(critical angle)
n2
(b) n1 > n2
If θi exceeds θc , the incident wave is totally reflected, and
the refracted wave becomes a nonuniform surface wave that
travels along the boundary between the two media. This wave
behavior is called total internal reflection.
No transmission
Example 8-4:
θr
θi
n1
θt = 90°
n2
Light Beam Passing through
a Slab
A dielectric slab with index of refraction n2 is surrounded by a
medium with index of refraction n1 , as shown in Fig. 8-11. If
θi < θc , show that the emerging beam is parallel to the incident
beam.
(c) n1 > n2 and θi = θc
θ1
n1
Figure 8-10
Snell’s laws state that θr = θi and
sin θt = (n1 /n2 ) sin θi . Refraction is (a) inward if n1 < n2
and (b) outward if n1 > n2 ; (c) the refraction angle is 90◦ if
n1 > n2 and θi is equal to or greater than the critical angle
θc = sin−1 (n2 /n1 ).
n2
θ2 θ2
n3 = n1
θ3 = θ1
referred to as more dense than another material if it has a
greater index of refraction.
At normal incidence (θi = 0), Eq. (8.31) gives θt = 0, as
expected. At oblique incidence θt < θi when n2 > n1 and
θt > θi when n2 < n1 .
◮ If a wave is incident on a more dense medium
(Fig. 8-10(a)), the transmitted wave refracts inwardly
(toward the z axis), and the opposite is true if a wave is
incident on a less dense medium (Fig. 8-10(b)). ◭
A case of particular interest is when θt = π /2, as shown in
Fig. 8-10(c); in this case, the refracted wave flows along the
surface and no energy is transmitted into medium 2. The value
of the angle of incidence θi corresponding to θt = π /2 is called
Figure 8-11 The exit angle θ3 is equal to the incidence
angle θ1 if the dielectric slab has parallel boundaries and is
surrounded by media with the same index of refraction on both
sides (Example 8-4).
Solution: At the slab’s upper surface, Snell’s law gives
n1
(8.33)
sin θ2 = sin θ1 .
n2
Similarly, at the slab’s lower surface,
n2
n2
sin θ3 = sin θ2 = sin θ2 .
n3
n1
Substituting Eq. (8.33) into Eq. (8.34) gives
n2
n1
sin θ3 =
sin θ1 = sin θ1 .
n1
n2
(8.34)
8-3 FIBER OPTICS
351
Hence, θ3 = θ1 . The slab displaces the beam’s position, but the
beam’s direction remains unchanged.
Exercise 8-4: In the visible part of the electromagnetic
spectrum, the index of refraction of water is 1.33. What is
the critical angle for light waves generated by an upwardlooking underwater light source?
Answer: θc = 48.8◦ . (See
EM
.)
Exercise 8-5: If the light source of Exercise 8-4 is situated
at a depth of 1 m below the water surface and if its beam
is isotropic (radiates in all directions), how large a circle
would it illuminate when observed from above?
Answer: Circle’s diameter = 2.28 m. (See
EM
critical angle θc for a wave in the fiber medium (with nf ) incident upon the cladding medium (with nc ). From Eq. (8.32a),
we have
nc
sin θc = .
(8.35)
nf
To meet the total reflection requirement θ3 ≥ θc , it is necessary
that sin θ3 ≥ nc /nf . The angle θ2 is the complement of angle θ3 ;
hence, cos θ2 = sin θ3 . The necessary condition therefore may
be written as
nc
(8.36)
cos θ2 ≥ .
nf
Moreover, θ2 is related to the incidence angle on the face of
the fiber, θi , by Snell’s law:
sin θ2 =
.)
8-3 Fiber Optics
By successive total internal reflections, as illustrated in
Fig. 8-12(a), light can be guided through thin dielectric rods
made of glass or transparent plastic known as optical fibers.
Because the light is confined to traveling within the rod, the
only loss in power is due to reflections at the sending and
receiving ends of the fiber and absorption by the fiber material
(because it is not a perfect dielectric). Optical fibers are useful
for the transmission of wide-band signals as well as many
imaging applications.
An optical fiber usually consists of a cylindrical fiber
core with an index of refraction nf , surrounded by another
cylinder of lower index of refraction, nc , called the cladding
(Fig. 8-12(b)). The cladding layer serves to optically isolate
the fiber when a large number of fibers are packed in close
proximity, thereby avoiding leakage of light from one fiber
into another. To ensure total internal reflection, the incident
angle θ3 in the fiber core must be equal to or greater than the
n0
sin θi ,
nf
(8.37)
where n0 is the index of refraction of the medium surrounding
the fiber (n0 = 1 for air and n0 = 1.33 if the fiber is in water)
or
"
#1/2
2
n0
2
cos θ2 = 1 −
.
(8.38)
sin θi
nf
Using Eq. (8.38) on the left-hand side of Eq. (8.36) and then
solving for sin θi gives
sin θi ≤
1 2
(n − n2c)1/2 .
n0 f
(8.39)
8-3.1 Maximum Acceptance Cone
The acceptance angle θa is defined as the maximum value
of θi for which the condition of total internal reflection remains
satisfied:
sin θa =
1 2
(n − n2c)1/2 .
n0 f
(8.40)
Acceptance cone
n0
nc
θ2
θi
θ3
nf
nc
n0
(a) Optical fiber
Fiber core
Cladding
θi
(b) Successive internal reflections
Figure 8-12 Waves can be guided along optical fibers as long as the reflection angles exceed the critical angle for total internal reflection.
352
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Cladding
Core
T
τi
τ
T
High-order mode
Low-order mode
Axial mode
Figure 8-13 Distortion of rectangular pulses caused by modal dispersion in optical fibers.
The angle θa is equal to half the angle of the acceptance cone
of the fiber. Any ray of light incident upon the face of the
core fiber at an incidence angle within the acceptance cone
can propagate down the core. This means that there can be a
large number of ray paths, called modes, by which light energy
can travel in the core. Rays characterized by large angles θi
travel longer paths than rays that propagate along the axis of
the fiber, as illustrated by the three modes shown in Fig. 8-13.
Consequently, different modes have different transit times
between the two ends of the fiber. This property of optical
fibers is called modal dispersion and has the undesirable effect
of changing the shape of pulses used for the transmission of
digital data. When a rectangular pulse of light incident upon
the face of the fiber gets broken up into many modes and the
different modes do not arrive at the other end of the fiber at the
same time, the pulse gets distorted—both in shape and length.
In the example shown in Fig. 8-13, the narrow rectangular
pulses at the input side of the optical fiber are of width τi
separated by a time duration T . After propagating through the
fiber core, modal dispersion causes the pulses to look more
like spread-out sine waves with spread-out temporal width τ .
If the output pulses spread out so much that τ > T , the output
signals will smear out, making it impossible to decipher the
transmitted message from the output signal. Hence, to ensure
that the transmitted pulses remain distinguishable at the output
side of the fiber, it is necessary that τ be shorter than T . As a
safety margin, it is common practice to design the transmission
system such that T ≥ 2τ .
The spread-out width τ is equal to the time delay ∆t between
the arrival of the slowest ray and the arrival of the fastest ray.
The slowest ray is the one traveling the longest distance and
corresponds to the ray incident upon the input face of the fiber
at the acceptance angle θa . From the geometry in Fig. 8-12(b)
and Eq. (8.36), this ray corresponds to cos θ2 = nc /nf . For an
optical fiber of length l, the length of the path traveled by such
a ray is
and its travel time in the fiber at velocity up = c/nf is
tmax =
ln2
lmax
= f .
up
cnc
(8.42)
The minimum time of travel is realized by the axial ray and is
given by
l
l
tmin =
= nf .
(8.43)
up
c
The total time delay is therefore
τ = ∆t = tmax − tmin =
lnf
c
nf − 1
nc
(s).
(8.44)
As we stated before, to retrieve the desired information from
the transmitted signals, it is advisable that T , which is the
interpulse period of the input train of pulses, be no shorter than
2τ . This in turn means that the maximum data rate (in bits per
second), or equivalently the number of pulses per second, that
can be transmitted through the fiber is limited to
fp =
1
1
cnc
=
=
T
2τ
2lnf (nf − nc )
Example 8-5:
(bits/s).
(8.45)
Transmission Data Rate
on Optical Fibers
A 1-km long optical fiber (in air) is made of a fiber core with
an index of refraction of 1.52 and a cladding with an index of
refraction of 1.49. Determine
(a) the acceptance angle θa and
(b) the maximum usable data rate of signals that can be
transmitted through the fiber.
Solution: (a) From Eq. (8.40),
lmax =
l
nf
=l ,
cos θ2
nc
(8.41)
sin θa =
1 2
(n − n2c)1/2 = [(1.52)2 − (1.49)2]1/2 = 0.3,
n0 f
8-4 WAVE REFLECTION AND TRANSMISSION AT OBLIQUE INCIDENCE
353
Module 8.2 Multimode Step-Index Optical Fiber Choose the indices on the fiber core and cladding and then observe
the zigzag pattern of the wave propagation inside the fiber.
which corresponds to θa = 17.5◦.
(b) From Eq. (8.45),
fp =
cnc
3 × 108 × 1.49
=
2lnf (nf − nc) 2 × 103 × 1.52(1.52 − 1.49)
= 4.9 (Mb/s).
8-3.2 Confined Acceptance Cone
The expression for the data rate given by Eq. (8.45) pertains to
the case where the light coupled to the fiber is confined to the
acceptance cone defined by the acceptance angle θa given by
Eq. (8.40). If the entrance cone is larger than the acceptance
cone, then some of the light will leak out of the fiber through
the cladding, which not only reduces the power carried by the
light but may also cause interference with signals carried by
neighboring fibers.
On the other hand, if the entrance cone can be made to
be smaller than the acceptance cone, the data rate can be
safely increased. Let us denote θi (max) as the maximum
angle of the entrance cone. Per Eq. (8.38), θ2 (max) is related
to θi (max) by
"
#1/2
2
n0
2
.
sin θi (max)
cos θ2 (max) = 1 −
nf
Repetition of the steps between Eqs. (8.41) and (8.45) leads to
c
cos θ2 (max)
(θi (max) < θa ).
fp =
2lnf 1 − cos θ2 (max)
Exercise 8-6: If the index of refraction of the cladding
material in Example 8-5 is increased to 1.50, what would
be the new maximum usable data rate?
Answer: 7.4 (Mb/s). (See
EM
.)
8-4 Wave Reflection and Transmission
at Oblique Incidence
In this section, we develop a rigorous theory of reflection
and refraction of plane waves obliquely incident upon planar
354
boundaries between different media. Our treatment parallels
that in Section 8-1 for the normal-incidence case and goes
beyond that in Section 8-2 on Snell’s laws, which yielded
information on only the angles of reflection and refraction.
For normal incidence, the reflection and transmission coefficients Γ and τ at a boundary between two media are independent of the polarization of the incident wave, because both
the electric and magnetic fields of a normally incident plane
wave are tangential to the boundary regardless of the wave
polarization. This is not the case for obliquely incident waves
traveling at an angle θi 6= 0 with respect to the normal to the
interface.
◮ The plane of incidence is defined as the plane containing the normal to the boundary and the direction of
propagation of the incident wave. ◭
A wave of arbitrary polarization may be described as the
superposition of two orthogonally polarized waves: one with
its electric field parallel to the plane of incidence (parallel
polarization) and the other with its electric field perpendicular to the plane of incidence (perpendicular polarization).
These two polarization configurations are shown in Fig. 8-14,
where the plane of incidence is coincident with the x–z plane.
Polarization with E perpendicular to the plane of incidence is
also called transverse electric (TE) polarization because E is
perpendicular to the plane of incidence and with E parallel
to the plane of incidence is called transverse magnetic (TM)
polarization because in that case it is the magnetic field that is
perpendicular to the plane of incidence.
For the general case of a wave with an arbitrary polarization,
it is common practice to decompose the incident wave (Ei , Hi )
into a perpendicularly polarized component (Ei⊥ , Hi⊥ ) and a
parallel polarized component (Eik , Hik ). Then, after determining the reflected waves (Er⊥ , Hr⊥ ) and (Erk , Hrk ) due to the two
incident components, the reflected waves are added together
to give the total reflected wave (Er , Hr ) corresponding to
the original incident wave. A similar process can be used to
determine the total transmitted wave (Et , Ht ).
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
x
ˆkr
r
HT
ETr
t
ET
ˆt
k
t
θr
θi
y
θt HT
z
ˆi
k
i
ET
i
HT
Medium 1
Medium 2
(ε1, μ1)
(ε2, μ2)
z=0
(a) Perpendicular polarization
x
ˆkr
r
E||
E||t
r
H||
ˆt
k
θr
θi
E||i
y
t
θt H||
z
ˆi
k
H||i
Medium 1
(ε1, μ1)
Medium 2
(ε2, μ2)
z=0
(b) Parallel polarization
Figure 8-14 The plane of incidence is the plane containing
the direction of wave travel, k̂i , and the surface normal to the
boundary. In the present case, the plane of incidence containing
k̂i and ẑ coincides with the plane of the paper. A wave is
(a) perpendicularly polarized when its electric field vector is
perpendicular to the plane of incidence and (b) parallel polarized
when its electric field vector lies in the plane of incidence.
8-4.1 Perpendicular Polarization
Figure 8-15 shows a perpendicularly polarized incident plane
wave propagating along the xi direction in dielectric medium 1.
e i points along the y direction, and
The electric field phasor E
⊥
e i is along the yi axis.
the associated magnetic field phasor H
⊥
e i points
e i are such that E
ei × H
e i and H
The directions of E
⊥
⊥
⊥
⊥
along the propagation direction x̂i . The electric and magnetic
fields of such a plane wave are given by
e i = ŷE i e− jk1 xi ,
E
⊥0
⊥
i
e i = ŷi E⊥0 e− jk1 xi ,
H
⊥
η1
(8.46a)
(8.46b)
8-4 WAVE REFLECTION AND TRANSMISSION AT OBLIQUE INCIDENCE
355
x
x
xr
yˆ r = xˆ cos θr + zˆ sin θr
HTrx
yr
θr
ETr
xr = x sin θr – z cos θr
Inset A
x xi
H Tr
θi z
z
HTrz
E
xr
–z θr
xi x
t
θt
θr
xi
Inset C
x xt
HTiz
HTix
θi
–x
HTtx
–x
HTtz
θt
H Tt
yt
z
yˆ t = –xˆ cos θt + zˆ sin θt
z
y
θi
E
xt
Tt
Inset B
xr x
Ti
xi = x sin θi + z cos θi
xt = x sin θt + z cos θt
θt z
z
H Ti
yˆ i = –xˆ cos θi + zˆ sin θi
yi
Medium 1 (ε1, μ1)
Medium 2 (ε2, μ2)
z=0
Figure 8-15 Perpendicularly polarized plane wave incident at an angle θi upon a planar boundary.
i
is the amplitude of the electric field phasor at xi = 0
where E⊥0
p
√
and k1 = ω µ1 ε1 and η1 = µ1 /ε1 are the wavenumber
and intrinsic impedance of medium 1. From Fig. 8-15, the
distance xi and the unit vector ŷi may be expressed in terms
of the (x, y, z) global coordinate system as
xi = x sin θi + z cos θi ,
(8.47a)
ŷi = −x̂ cos θi + ẑsin θi .
(8.47b)
Substituting Eqs. (8.47a) and (8.47b) into Eqs. (8.46a) and
(8.46b) gives
Incident Wave
e i = ŷE i e− jk1 (x sin θi +z cos θi ) ,
E
⊥0
⊥
= (x̂ cos θr + ẑsin θr )
r
E⊥0
e− jk1 (x sin θr −z cos θr ) ,
η1
(8.48a)
With the aid of the directional relationships given in Fig. 8-15
for the reflected and transmitted waves, these fields are given
by the following:
Reflected Wave
(8.49a)
(8.49b)
Transmitted Wave
e t = ŷE t e− jk2 xt = ŷE t e− jk2 (x sin θt +z cos θt ) ,
E
⊥0
⊥0
⊥
(8.49c)
t
e t = ŷt E⊥0 e− jk2 xt
H
⊥
η2
= (−x̂ cos θt + ẑ sin θt )
i
e i = (−x̂ cos θi + ẑsin θi ) E⊥0 e− jk1 (x sin θi +z cos θi ) . (8.48b)
H
⊥
η1
e r = ŷE r e− jk1 xr = ŷE r e− jk1 (x sin θr −z cos θr ) ,
E
⊥0
⊥0
⊥
r
e r = ŷr E⊥0 e− jk1 xr
H
⊥
η1
t
E⊥0
e− jk2 (x sin θt +z cos θt ) , (8.49d)
η2
where θr and θt are the reflection and transmission angles
shown in Fig. 8-15 and k2 and η2 are the wavenumber and
intrinsic impedance of medium 2. Our goal is to describe the
reflected and transmitted fields in terms of the parameters
that characterize the incident wave, namely the incidence
i
. The four expressions given
angle θi and the amplitude E⊥0
r
by Eqs. (8.49a) through (8.49d) contain four unknowns: E⊥0
,
t
E⊥0 , θr , and θt . Even though angles θr and θt are related to θi
by Snell’s laws (Eqs. (8.28a) and (8.28b)), here we choose to
356
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
treat them as unknown for the time being because we intend
to show that Snell’s laws also can be derived by applying
field boundary conditions at z = 0. The total electric field in
medium 1 is the sum of the incident and reflected electric
e r ; a similar statement holds true for the
ei + E
e1 = E
fields: E
⊥
⊥
⊥
e r . Boundary
ei + H
e1 = H
total magnetic field in medium 1: H
⊥
⊥
⊥
e and H
e
conditions state that the tangential components of E
each must be continuous across the boundary between the two
media. Field components tangential to the boundary extend
along x̂ and ŷ. Since the electric fields in media 1 and 2 have ŷ
e is
components only, the boundary condition for E
i
r
(Ee⊥y
+ Ee⊥y
)
z=0
t
= Ee⊥y
z=0
.
(8.50)
Upon using Eqs. (8.48a), (8.49a), and (8.49c) in Eq. (8.50) and
then setting z = 0, we have
i
r
t
E⊥0
e− jk1 x sin θi + E⊥0
e− jk1 x sin θr = E⊥0
e− jk2 x sin θt .
(8.51)
Since the magnetic fields in media 1 and 2 have no ŷ compoe is
nents, the boundary condition for H
i
r
e⊥x
e⊥x
(H
+H
)
z=0
t
e⊥x
=H
z=0
,
(8.52)
i
E⊥0
Er
cos θi e− jk1 x sin θi + ⊥0 cos θr e− jk1 x sin θr
η1
η1
=−
t
E⊥0
cos θt e− jk2 x sin θt .
η2
(8.53)
To satisfy Eqs. (8.51) and (8.53) for all possible values of x
(i.e., all along the boundary), it follows that the arguments of
all three exponentials must be equal. That is,
k1 sin θi = k1 sin θr = k2 sin θt ,
(8.54)
which is known as the phase-matching condition. The first
equality in Eq. (8.54) leads to
θr = θi ,
(Snell’s law of reflection)
(8.55)
while the second equality leads to
√
ω µ1 ε1
k1
n1
sin θt
=
= .
= √
sin θi
k2
ω µ2 ε2
n2
(Snell’s law of refraction)
i
r
t
E⊥0
+ E⊥0
= E⊥0
,
cos θt t
cos θi
i
r
(−E⊥0
+ E⊥0
)=−
E⊥0 .
η1
η2
(8.57a)
(8.57b)
These two equations can be solved simultaneously to yield
the following expressions for the reflection and transmission
coefficients in the perpendicular polarization case:
Γ⊥ =
τ⊥ =
r
E⊥0
η2 cos θi − η1 cos θt
,
=
i
E⊥0 η2 cos θi + η1 cos θt
t
E⊥0
2η2 cos θi
=
.
i
E⊥0 η2 cos θi + η1 cos θt
(8.58a)
(8.58b)
These two coefficients are known formally as the Fresnel
reflection and transmission coefficients for perpendicular
polarization and are related by
or
−
The results expressed by Eqs. (8.55) and (8.56) are identical
with those derived previously in Section 8-2 through consideration of the ray path traversed by the incident, reflected, and
transmitted wavefronts.
In view of Eq. (8.54), the boundary conditions given by
Eqs. (8.51) and (8.53) reduce to
τ⊥ = 1 + Γ⊥.
(8.59)
If medium 2 is a perfect conductor (η2 = 0), Eqs. (8.58a and b)
reduce to Γ⊥ = −1 and τ⊥ = 0, respectively, which means
that the incident wave is totally reflected by the conducting
medium.
For nonmagnetic dielectrics with µ1 = µ2 = µ0 and with the
help of Eq. (8.56), the expression for Γ⊥ can be written as
q
(ε2 /ε1 ) − sin2 θi
q
Γ⊥ =
cos θi + (ε2 /ε1 ) − sin2 θi
cos θi −
(8.60)
(for µ1 = µ2 ).
(8.56)
Since (ε2 /ε1 ) = (n2 /n1 )2 , this expression also can be written
in terms of the indices of refraction n1 and n2 .
8-4 WAVE REFLECTION AND TRANSMISSION AT OBLIQUE INCIDENCE
Example 8-6:
The corresponding wavenumber in medium 2 is
Wave Incident Obliquely
on a Soil Surface
k2 =
Using the coordinate system of Fig. 8-15, a plane wave
radiated by a distant antenna is incident in air upon a plane
soil surface located at z = 0. The electric field of the incident
wave is given by
Ei = ŷ100 cos(ω t − π x − 1.73π z)
(V/m),
e i = ŷ100e− jπ x− j1.73π z = ŷ100e− jk1xi
E
(V/m),
(8.62)
where xi is the axis along which the wave is traveling and
k1 xi = π x + 1.73π z.
(8.63)
k1 xi = k1 x sin θi + k1 z cos θi .
θt = 14.5◦.
With ε1 = ε0 and ε2 = εr2 ε0 = 4ε0 , the reflection and transmission coefficients for perpendicular polarization are determined
with the help of Eqs. (8.59) and (8.60),
q
cos θi − (ε2 /ε1 ) − sin2 θi
q
= −0.38,
Γ⊥ =
cos θi + (ε2 /ε1 ) − sin2 θi
τ⊥ = 1 + Γ⊥ = 0.62.
i
Using Eqs. (8.48a) and (8.49a) with E⊥0
= 100 V/m and θi =
θr , the total electric field in medium 1 is
er
ei + E
e1 = E
E
⊥
⊥
⊥
i
− jk1 (x sin θi +z cos θi )
i
+ ŷΓE⊥0
e− jk1 (x sin θi −z cos θi )
= ŷ100e− j(π x+1.73π z) − ŷ38e− j(π x−1.73π z),
and the corresponding instantaneous electric field in medium 1
is
h
i
e 1 e jω t
E1⊥ (x, z,t) = Re E
⊥
Hence,
k1 sin θi = π ,
k1 cos θi = 1.73π ,
= ŷ[100 cos(ω t − π x − 1.73π z)
− 38 cos(ω t − π x + 1.73π z)]
(rad/m),
The wavelength in medium 1 (air) is
λ1 =
(8.64)
k1
2π
sin 30◦ = 0.25
sin θi =
k2
4π
or
= ŷE⊥0 e
Using Eq. (8.47a), we have
(rad/m).
(b) Given that θi = 30◦ , the transmission angle θt is obtained
with the help of Eq. (8.56):
sin θt =
Solution: (a) We begin by converting Eq. (8.61) into phasor
form, akin to the expression given by Eq. (8.46a):
2π
= 4π
λ2
e i is along ŷ, it is perpendicularly polarized (ŷ is
Since E
perpendicular to the plane of incidence containing the surface
normal ẑ and the propagation direction x̂i ).
(8.61)
and the soil medium may be assumed to be a lossless dielectric
with a relative permittivity of 4.
(a) Determine k1 , k2 , and the incidence angle θi .
(b) Obtain expressions for the total electric fields in air and
in the soil.
(c) Determine the average power density carried by the wave
traveling in soil.
which together give
q
k1 = π 2 + (1.73π )2 = 2π
π = 30◦ .
θi = tan−1
1.73π
357
2π
= 1 m,
k1
and the wavelength in medium 2 (soil) is
λ1
1
λ2 = √ = √ = 0.5 m.
εr2
4
(V/m).
i
t
gives
In medium 2, using Eq. (8.49c) with E⊥0
= τ⊥ E⊥0
e t = ŷτ E i e− jk2 (x sin θt +z cos θt ) = ŷ62e− j(π x+3.87π z)
E
⊥0
⊥
and, correspondingly,
i
h
e t e jω t
Et⊥ (x, z,t) = Re E
⊥
= ŷ62 cos(ω t − π x − 3.87π z)
(V/m).
√
√
(c) In medium 2, η2 = η0 / εr2 ≈ 120π / 4 = 60π (Ω), and
the average power density carried by the wave is
t
Sav
=
t 2
|E⊥0
|
(62)2
=
= 10.2
2η2
2 × 60π
(W/m2 ).
358
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Technology Brief 16:
Bar-Code Readers
A bar code consists of a sequence of parallel bars of
certain widths, usually printed in black against a white
background, configured to represent a particular binary
code of information about a product and its manufacturer. Laser scanners can read the code and transfer the
information to a computer, a cash register, or a display
screen. For both stationary scanners built into checkout
counters at grocery stores and handheld units that can
be pointed at the bar-coded object like a gun, the basic
operation of a bar-code reader is the same.
Bar code
Electrical signal
Digital code
Figure TF16-2 Bar code contained in reflected laser beam.
Basic Operation
The scanner uses a laser beam of light pointed at
a multifaceted rotating mirror, spinning at a high
speed on the order of 6,000 revolutions per minute
(Fig. TF16-1). The rotating mirror creates a fan beam
to illuminate the bar code on the object. Moreover,
by exposing the laser light to its many facets, it
deflects the beam into many different directions,
allowing the object to be scanned over a wide range
of positions and orientations. The goal is to have
one of those directions be such that the beam reflected
by the bar code ends up traveling in the direction of,
and captured by, the light detector (sensor), which then
reads the coded sequence (white bars reflect laser light
and black ones do not) and converts it into a binary
sequence of ones and zeros (Fig. TF16-2). To eliminate
interference by ambient light, a glass filter is used as
shown in Fig. TF16-1 to block out all light except for a
narrow wavelength band centered at the wavelength of
the laser light.
Bar code
Central store
computer
Glass filter
Cash register
Sensor
Rotating mirror
(6,000 rpm)
Figure TF16-1 Elements of a bar-code reader.
8-4 WAVE REFLECTION AND TRANSMISSION AT OBLIQUE INCIDENCE
8-4.2 Parallel Polarization
359
Reflected Wave
If we interchange the roles played by E and H in the perpendicular polarization scenario covered in the preceding subsection,
× H must point in
while keeping in mind the requirement that E×
the direction of propagation for each of the incident, reflected,
and transmitted waves, we end up with the parallel polarization
scenario shown in Fig. 8-16. Now the electric fields lie in the
plane of incidence, while the associated magnetic fields are
perpendicular to the plane of incidence. With reference to the
directions indicated in Fig. 8-16, the fields of the incident,
reflected, and transmitted waves are given by
e r = ŷr E r e− jk1 xr
E
k0
k
r − jk1 (x sin θr −z cos θr )
e
= (x̂ cos θr + ẑsin θr )Ek0
,
e r = −ŷ
H
k
r
Ek0
η1
e− jk1 xr = −ŷ
r
Ek0
η1
e− jk1 (x sin θr −z cos θr ) , (8.65d)
Transmitted Wave
e t = ŷt E t e− jk2 xt
E
k0
k
t − jk2 (x sin θt +z cos θt )
= (x̂ cos θt − ẑsin θt )Ek0
,
e
Incident Wave
e t = ŷ
H
k
e i = ŷi E i e− jk1 xi
E
k0
k
i − jk1 (x sin θi +z cos θi )
e
= (x̂ cos θi − ẑsin θi )Ek0
,
e i = ŷ
H
k
i
Ek0
η1
e− jk1 xi = ŷ
i
Ek0
η1
e− jk1 (x sin θi +z cos θi ) ,
(8.65a)
(8.65b)
t
Ek0
η2
x
τk =
yr
E||rx
H||r
θr
E||r
E||rz
z
t
E||t θt E||x
xr
θi
–z
E||iz
E||tz
xt
η2
e− jk2 (x sin θt +z cos θt ) .
(8.65f)
r
Ek0
Ei
=
η2 cos θt − η1 cos θi
,
η2 cos θt + η1 cos θi
(8.66a)
=
2η2 cos θi
.
η2 cos θt + η1 cos θi
(8.66b)
k0
t
Ek0
Ei
k0
The preceding expressions can be shown to yield the relation
xt
y
xi
i
H||
τk = (1 + Γk)
H||t
θt
θr
E||i
i
θi E||x
t
Ek0
yt
–z
yi
e− jk2 xt = ŷ
(8.65e)
e and H
e in both
By matching the tangential components of E
media at z = 0, we again obtain the relations defining Snell’s
laws, as well as the following expressions for the Fresnel
reflection and transmission coefficients for parallel polarization:
Γk =
xr
(8.65c)
z
xi = x sin θi + z cos θi
yˆ i = xˆ cos θi – zˆ sin θi
xr = x sin θr – z cos θr
yˆ r = xˆ cos θr + zˆ sin θr
xt = x sin θt + z cos θt
yˆ t = xˆ cos θt – zˆ sin θt
Medium 1 (ε1, μ1)
Medium 2 (ε2, μ2)
Figure 8-16 Parallel-polarized plane wave incident at an
angle θi upon a planar boundary.
cos θi
.
cos θt
(8.67)
We noted earlier in connection with the perpendicularpolarization case that, when the second medium is a perfect
conductor with η2 = 0, the incident wave gets totally reflected
at the boundary. The same is true for the parallel polarization
case; setting η2 = 0 in Eqs. (8.66a and b) gives Γk = −1 and
τk = 0.
For nonmagnetic materials, Eq. (8.66a) becomes
q
−(ε2 /ε1 ) cos θi + (ε2 /ε1 ) − sin2 θi
q
Γk =
(ε2 /ε1 ) cos θi + (ε2 /ε1 ) − sin2 θi
(for µ1 = µ2 ).
(8.68)
360
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Perpendicular Polarization
1
0.8
For perpendicular polarization, the Brewster angle θB⊥ can be
obtained by setting the numerator of the expression for Γ⊥ ,
given by Eq. (8.58a), equal to zero. This happens when
Water
(εr = 81)
|ГT|
η2 cos θi = η1 cos θt .
|ГT| 0.6 Wet soil
(εr = 25)
or
|Г| ||
0.4
Dry soil
(εr = 3)
By (1) squaring both sides of Eq. (8.69), (2) using Eq. (8.56),
(3) solving for θi , and then denoting θi as θB⊥ , we obtain
s
1 − (µ1ε2 /µ2 ε1 )
.
(8.70)
sin θB⊥ =
1 − (µ1/ µ2 )2
|Г|||
0.2
0
(8.69)
Because the denominator of Eq. (8.70) goes to zero when
µ1 = µ2 , θB⊥ does not exist for nonmagnetic materials.
10
20
30
40
50
60
70
80
90
(θB dry soil) (θB wet soil) (θB water)
Incidence angle θi (degrees)
Figure 8-17 Plots for |Γ⊥ | and |Γk | as a function of θi for a
dry soil surface, a wet soil surface, and a water surface. For each
surface, |Γk | = 0 at the Brewster angle.
To illustrate the angular variations of the magnitudes of Γ⊥
and Γk , Fig. 8-17 shows plots for waves incident in air onto
three different types of dielectric surfaces: dry soil (εr = 3),
wet soil (εr = 25), and water (εr = 81). For each of the surfaces,
(1) Γ⊥ = Γk at normal incidence (θi = 0), as expected; (2)
|Γ⊥ | = |Γk | = 1 at grazing incidence (θi = 90◦); and (3) Γk
goes to zero at an angle called the Brewster angle in Fig. 8-17.
Had the materials been magnetic too (µ1 6= µ2 ), it would have
been possible for Γ⊥ to vanish at some angle as well. However,
for nonmagnetic materials, the Brewster angle exists only for
parallel polarization, and its value depends on the ratio (ε2 /ε1 ),
as we see shortly.
◮ At the Brewster angle, the parallel-polarized component of the incident wave is totally transmitted into
medium 2. ◭
8-4.3 Brewster Angle
The Brewster angle θB is defined as the incidence angle θi at
which the Fresnel reflection coefficient Γ = 0.
Parallel Polarization
For parallel polarization, the Brewster angle θBk at which
Γk = 0 can be found by setting the numerator of Γk in
Eq. (8.66a) equal to zero. The result is identical to Eq. (8.70),
but with µ and ε interchanged. That is,
s
1 − (ε1 µ2 /ε2 µ1 )
.
(8.71)
sin θBk =
1 − (ε1 /ε2 )2
For nonmagnetic materials,
−1
θBk = sin
−1
= tan
s
r
1
1 + (ε1 /ε2 )
ε2
ε1
(for µ1 = µ2 ).
(8.72)
The Brewster angle is also called the polarizing angle. This
is because, if a wave composed of both perpendicular and
parallel polarization components is incident upon a nonmagnetic surface at the Brewster angle θBk , the parallel polarized
component is totally transmitted into the second medium, and
only the perpendicularly polarized component is reflected by
the surface. Natural light, including sunlight and light generated by most manufactured sources, is unpolarized because
the direction of the electric field of the light waves varies
randomly in angle over the plane perpendicular to the direction
of propagation. Thus, on average, half of the intensity of
natural light is perpendicularly polarized and the other half is
parallel polarized. When unpolarized light is incident upon a
surface at the Brewster angle, the reflected wave is strictly perpendicularly polarized. Hence, the surface acts as a polarizer.
8-5 REFLECTIVITY AND TRANSMISSIVITY
Pr
Medium 1 (ε1, μ1)
s
co
co
s
θi
Pi
A
Can total internal reflection
take place for a wave incident from medium 1 (with n1 )
onto medium 2 (with n2 ) when n2 > n1 ?
θr
A
Concept Question 8-4:
361
What is the difference between
the boundary conditions applied in Section 8-1.1 for
normal incidence and those applied in Section 8-4.1 for
oblique incidence with perpendicular polarization?
Concept Question 8-5:
θi θr
A
Why is the Brewster angle also
called the polarizing angle?
Concept Question 8-6:
θt
At the boundary, the vector
sum of the tangential components of the incident and
reflected electric fields has to equal the tangential component of the transmitted electric field. For εr1 = 1 and
εr2 = 16, determine the Brewster angle and then verify the
validity of the preceding statement by sketching to scale
the tangential components of the three electric fields at the
Brewster angle.
Concept Question 8-7:
Medium 2 (ε2, μ2)
os
Ac
θt
Pt
Figure 8-18 Reflection and transmission of an incident circular beam illuminating a spot of size A on the interface.
Exercise 8-7: A wave in air is incident upon a soil surface
at θi = 50◦ . If soil has εr = 4 and µr = 1, determine Γ⊥ ,
τ⊥ , Γk , and τk .
Answer: Γ⊥ = −0.48, τ⊥ = 0.52, Γk = −0.16, τk = 0.58.
(See
EM
.)
Exercise 8-8: Determine the Brewster angle for the
boundary of Exercise 8.7.
Answer: θB = 63.4◦ . (See
EM
energy incident upon the boundary between two contiguous,
lossless media. The area of the spot illuminated by the beam
is A, and the incident, reflected, and transmitted beams have
r
t
i
, E⊥0
, and E⊥0
, respectively. The
electric-field amplitudes E⊥0
average power densities carried by the incident, reflected, and
transmitted beams are
.)
i
S⊥
=
i 2
|E⊥0
|
,
2η1
(8.73a)
mitted electric and magnetic fields given by Eqs. (8.65a)
through (8.65f) all have the same exponential phase function along the x direction.
r
S⊥
=
r 2
|E⊥0
|
,
2η1
(8.73b)
Answer: With the help of Eqs. (8.55) and (8.56), all six
t
=
S⊥
t 2
|E⊥0
|
,
2η2
(8.73c)
Exercise 8-9: Show that the incident, reflected, and trans-
fields are shown to vary as e− jk1 x sin θi . (See
EM
.)
8-5 Reflectivity and Transmissivity
The reflection and transmission coefficients derived earlier are
ratios of the reflected and transmitted electric field amplitudes
to the amplitude of the incident electric field. We now examine power ratios, starting with the perpendicular polarization
case. Figure 8-18 shows a circular beam of electromagnetic
where η1 and η2 are the intrinsic impedances of media 1
and 2, respectively. The cross-sectional areas of the incident,
reflected, and transmitted beams are
Ai = A cos θi ,
(8.74a)
Ar = A cos θr ,
(8.74b)
At = A cos θt ,
(8.74c)
362
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
and the corresponding average powers carried by the beams
are
(8.75a)
In fact, in many cases, the transmitted electric field is larger
than the incident electric field. Conservation of power requires
that the incident power equals the sum of the reflected and
transmitted powers. That is, for perpendicular polarization,
|E r |2
P⊥ = S⊥ Ar = ⊥0 A cos θr ,
2η1
(8.75b)
P⊥i = P⊥r + P⊥t ,
t 2
|E⊥0
|
A cos θt .
2η2
(8.75c)
i
P⊥i = S⊥
Ai =
r
i 2
|E⊥0
|
A cos θi ,
2η1
r
t
P⊥t = S⊥
At =
or
The reflectivity R (also called reflectance in optics) is defined
as the ratio of the reflected to the incident power. The reflectivity for perpendicular polarization is then
R⊥ =
r
r 2
E⊥0
|E⊥0
| cos θr
P⊥r
=
=
i |2 cos θ
i
P⊥i
|E⊥0
E⊥0
i
i 2
|E t |2
|E⊥0
|
|E r |2
A cos θi = ⊥0 A cos θr + ⊥0 A cos θt . (8.81)
2η1
2η1
2η2
Use of Eqs. (8.76), (8.79a), and (8.79b) leads to
2
,
(8.76)
where we used the fact that θr = θi in accordance with Snell’s
law of reflection. The ratio of the reflected to incident electric
r
i
field amplitudes, |E⊥0
/E⊥0
|, is equal to the magnitude of the
reflection coefficient Γ⊥ . Hence,
R⊥ = |Γ⊥ |2 ,
R⊥ + T⊥ = 1,
(8.82a)
Rk + Tk = 1,
(8.82b)
or
|Γ⊥ |2 + |τ⊥ |2
η1 cos θt
η2 cos θi
= 1,
(8.83a)
2
η1 cos θt
η2 cos θi
= 1.
(8.83b)
(8.77)
2
|Γk | + |τk |
and similarly, for parallel polarization
Pki
2
= |Γk | .
(8.78)
The transmissivity T (or transmittance in optics) is defined as
the ratio of the transmitted power to incident power:
t 2
|E⊥0
| η1 A cos θt
P⊥t
=
i
i
P⊥
|E⊥0 |2 η2 A cos θi
2 η1 cos θt
,
= |τ⊥ |
η2 cos θi
Pkt
2 η1 cos θt
.
Tk = i = |τk |
η2 cos θi
Pk
Figure 8-19 shows plots for (Rk , Tk ) as a function of θi for an
air–glass interface. Note that the sum of Rk and Tk is always
equal to 1, as mandated by Eq. (8.82b). We also note that, at
the Brewster angle θB , Rk = 0 and Tk = 1.
Table 8-2 provides a summary of the general expressions
for Γ, τ , R, and T for both normal and oblique incidence.
1
T⊥ =
(8.79a)
T| |
Reflectivity R| | and
transmissivity T| |
Rk =
Pkr
(8.80)
θi
Air
Glass
0.5
(8.79b)
0
◮ The incident, reflected, and transmitted waves do not
have to obey any such laws as conservation of electric
field, conservation of magnetic field, or conservation of
power density, but they do have to obey the law of
conservation of power. ◭
n = 1.5
θB
0
60
30
R| |
90
θi (degrees)
Figure 8-19 Angular plots for (Rk , Tk ) for an air–glass interface.
8-5 REFLECTIVITY AND TRANSMISSIVITY
363
Table 8-2 Expressions for Γ, τ , R, and T for wave incidence from a medium with intrinsic impedance η1 onto a medium with
intrinsic impedance η2 . Angles θi and θt are the angles of incidence and transmission, respectively.
Normal Incidence
θi = θt = 0
Property
Perpendicular
Polarization
Parallel
Polarization
Reflection coefficient
Γ=
η2 − η1
η2 + η1
Γ⊥ =
η2 cos θi − η1 cos θt
η2 cos θi + η1 cos θt
Γk =
η2 cos θt − η1 cos θi
η2 cos θt + η1 cos θi
Transmission coefficient
τ=
2η2
η2 + η1
τ⊥ =
2η2 cos θi
η2 cos θi + η1 cos θt
τk =
2η2 cos θi
η2 cos θt + η1 cos θi
Relation of Γ to τ
τ = 1+Γ
τ⊥ = 1 + Γ⊥
τk = (1 + Γk )
Reflectivity
R = |Γ|2
R⊥ = |Γ⊥ |2
Rk = |Γk |2
Transmissivity
T = |τ |2
Relation of R to T
Notes: (1) sin θt =
p
η1
η2
T = 1−R
µ1 ε1 /µ2 ε2 sin θi ; (2) η1 =
T⊥ = |τ⊥ |2
p
η1 cos θt
η2 cos θi
T⊥ = 1 − R⊥
µ1 /ε1 ; (3) η2 =
p
Tk = |τk |2
cos θi
cos θt
η1 cos θt
η2 cos θi
Tk = 1 − Rk
µ2 /ε2 ; (4) for nonmagnetic media, η2 /η1 = n1 /n2 .
Module 8.3 Oblique Incidence Upon specifying the frequency, polarization and incidence angle of a plane wave incident
upon a planar boundary between two lossless media, this module displays vector information and plots of the reflection and
transmission coefficients as a function of the incidence angle.
364
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Module 8.4 Oblique Incidence in Lossy Medium This module extends the capabilities of Module 8.1 to situations in
which medium 2 is lossy.
Example 8-7:
Beam of Light
A 5 W beam of light with circular cross section is incident
in air upon the plane boundary of a dielectric medium with
index of refraction of 5. If the angle of incidence is 60◦ and the
incident wave is parallel polarized, determine the transmission
angle and the powers contained in the reflected and transmitted
beams.
Solution: From Eq. (8.56),
sin θt =
or
n1
1
sin θi = sin 60◦ = 0.17
n2
5
θt = 10◦ .
With ε2 /ε1 = n22 /n21 = (5)2 = 25, the reflection coefficient for
parallel polarization follows from Eq. (8.68) as
q
−(ε2 /ε1 ) cos θi + (ε2 /ε1 ) − sin2 θi
q
Γk =
(ε2 /ε1 ) cos θi + (ε2 /ε1 ) − sin2 θi
p
−25 cos60◦ + 25 − sin2 60◦
p
=
= −0.435.
25 cos60◦ + 25 − sin2 60◦
The reflected and transmitted powers therefore are
Pkr = Pki |Γk |2 = 5(0.435)2 = 0.95 W,
Pkt = Pki − Pkr = 5 − 0.95 = 4.05 W.
8-6 Waveguides
Earlier in Chapter 2, we considered two families of
transmission lines, namely those that support transverseelectromagnetic (TEM) modes, and those that do not. Transmission lines belonging to the TEM family (Fig. 2-4), including coaxial, two-wire, and parallel-plate lines, support E and H
fields that are orthogonal to the direction of propagation. Fields
supported by lines in the other group—often called higherorder transmission lines—may have E or H orthogonal to the
direction of propagation k̂, but not both simultaneously. Thus,
at least one component of E or H is along k̂.
◮ If E is transverse to k̂ but H is not, we call it a
transverse electric (TE) mode, and if H is transverse
to k̂ but E is not, we call it a transverse magnetic (TM)
mode. ◭
8-6 WAVEGUIDES
365
n0
nc
θi
nf
θ 2 θ3
nc
n0
Waveguide
y
Fiber core
Cladding
x
θi
a
b
(a) Optical fiber
Probe
0
Metal
z
Hollow or
dielectric-filled
Coaxial line
(a) Coax-to-waveguide coupler
Electric field
y=b
EM wave
(b) Circular waveguide
Metal
Hollow or
dielectric-filled
y=0
z
(b) Cross-sectional view at x = a/2
Figure 8-21 The inner conductor of a coaxial cable can excite
an EM wave in the waveguide.
(c) Rectangular waveguide
Figure 8-20 Wave travel by successive reflections in (a) an
optical fiber, (b) a circular metal waveguide, and (c) a rectangular
metal waveguide.
Among all higher-order transmission lines, the two most
commonly used are optical fiber and the metal waveguide.
As noted in Section 8-3, a wave is guided along an optical
fiber through successive zigzags by taking advantage of the
total internal reflection at the boundary between the (inner)
core and the (outer) cladding (Fig. 8-20(a)). Another way to
achieve internal reflection at the core’s boundary is to have
its surface coated by a conducting material. Under the proper
conditions (we shall elaborate later) a wave excited in the
interior of a hollow conducting pipe, such as the circular
or rectangular waveguides shown in Figs. 8-20(b) and (c),
undergoes a process similar to that of successive internal
reflection in an optical fiber, resulting in propagation down the
pipe. Most waveguide applications call for air-filled guides, but
in some cases, the waveguide may be filled with a dielectric
material used to alter its propagation velocity or impedance, or
it may be vacuum-pumped to eliminate air molecules in order
to prevent voltage breakdown, thereby increasing its powerhandling capabilities.
Figure 8-21 illustrates how a coaxial cable can be connected
to a rectangular waveguide. With its outer conductor connected
to the metallic waveguide enclosure, the coaxial cable’s inner
conductor protrudes through a tiny hole into the waveguide’s
interior (without touching the conducting surface). Timevarying electric field lines extending between the protruding
inner conductor and the inside surface of the guide provide
the excitation necessary to transfer a signal from the coaxial
366
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
line to the guide. Conversely, the center conductor can act like
a probe, coupling a signal from the waveguide to the coaxial
cable.
For guided transmission at frequencies below 30 GHz, the
coaxial cable is by far the most widely used transmission line.
At higher frequencies, however, the coaxial cable has a number
of limitations: (a) in order for it to propagate only TEM modes,
the cable’s inner and outer conductors have to be reduced
in size to satisfy a certain size-to-wavelength requirement,
making it more difficult to fabricate, (b) the smaller cross section reduces the cable’s power-handling capacity (limited by
dielectric breakdown), and (c) the attenuation due to dielectric
losses increases with frequency. For all of these reasons, metal
waveguides have been used as an alternative to coaxial lines
for many radar and communication applications that operate
at frequencies in the 5–100 GHz range, particularly those requiring the transmission of high levels of radio-frequency (RF)
power. Even though waveguides with circular and elliptical
cross sections have been used in some microwave systems, the
rectangular shape has been the more prevalent geometry.
8-7 General Relations for E and H
The purpose of the next two sections is to derive expressions
for E and H for the TE and TM modes in a rectangular waveguide and to examine their wave properties. We choose the
coordinate system shown in Fig. 8-22, where the propagation
occurs along ẑ. For TE modes, the electric field is transverse to
the direction of propagation. Hence, E may have components
along x̂ and ŷ but not along ẑ. In contrast, H has a ẑ-directed
component and may have components along either x̂ or ŷ or
along both. The converse is true for TM modes.
Our solution procedure consists of four steps:
(1) Maxwell’s equations are manipulated to develop general
expressions for the phasor-domain transverse field comex , and H
ey in terms of Eez and H
ez . When
ponents Eex , Eey , H
z
y
b
x
a
0
Figure 8-22 Waveguide coordinate system.
specialized to the TE case, these expressions become
ez only, and the converse is true for the TM
functions of H
case.
(2) The homogeneous wave equations given by Eqs. (7.15)
and (7.16) are solved to obtain valid solutions for Eez (TM
ez (TE case) in a waveguide.
case) and H
(3) The expressions derived in step 1 are then used to find Eex ,
ex , and H
ey .
Eey , H
(4) The solutions obtained in step 3 are analyzed to determine
the phase velocity and other properties of the TE and TM
waves.
The intent of the present section is to realize the stated goals
of step 1. We begin with a general form for the E and H fields
in the phasor domain:
e = x̂ Eex + ŷ Eey + ẑ Eez ,
E
(8.84a)
e = x̂ H
ex + ŷ H
ey + ẑ H
ez .
H
(8.84b)
Eex (x, y, z) = eex (x, y) e− jβ z ,
(8.85)
e = (x̂ eex + ŷ eey + ẑ eez )e− jβ z ,
E
(8.86a)
e and H
e may depend on
In general, all six components of E
(x, y, z), and while we do not yet know how they functionally
e and H
e
depend on (x, y), our prior experience suggests that E
of a wave traveling along the +z direction should exhibit a
dependence on z of the form e− jβ z, where β is a yet-to-be
determined phase constant. Hence, we adopt the form
where eex (x, y) describes the dependence of Eex (x, y, z) on (x, y)
only. The form of Eq. (8.85) can be used for all other compoe and H
e as well. Thus,
nents of E
e = (x̂ e
H
hx + ŷ e
hy + ẑ e
hz )e− jβ z .
(8.86b)
e = − jω µ H,
e
∇×E
(8.87a)
e
The notation is intended to clarify that, in contrast to Ee and H,
e
which vary with (x, y, z), the lower case ee and h vary with (x, y)
only.
In a lossless, source-free medium (such as the inside of a
waveguide) characterized by permittivity ε and permeability µ
(and conductivity σ = 0), Maxwell’s curl equations are given
by Eqs. (7.2b and d) with J = 0,
e
e = jωε E.
∇×H
(8.87b)
Upon inserting Eqs. (8.86a and b) into Eqs. (8.87a and b) and
recalling that each of the curl equations actually consists of
8-8 TM MODES IN RECTANGULAR WAVEGUIDE
three separate equations—one for each of the unit vectors x̂, ŷ,
and ẑ, we obtain the following relationships:
∂ eez
+ jβ eey = − jω µ e
hx ,
∂y
∂ eez
hy ,
= − jω µ e
∂x
∂ eey ∂ eex
−
= − jω µ e
hz ,
∂x
∂y
− jβ eex −
∂e
hz
+ jβ e
hy = jωε eex ,
∂y
∂e
hz
= jωε eey ,
∂x
hy ∂ e
∂e
hx
−
= jωε eez .
∂x
∂y
− jβ e
hx −
(8.88a)
(8.88b)
(8.88c)
−j
Eex = 2
kc
β
ez
∂ Eez
∂H
+ ωµ
∂x
∂y
(8.88e)
−β
In the preceding section, we developed expressions for Eex , Eey ,
ez = 0 for the TM
ex , and H
ey in terms of Eez and H
ez . Since H
H
mode, our task reduces to obtaining a valid solution for Eez .
e
Our starting point is the homogeneous wave equation for E.
For a lossless medium characterized by an unbounded-medium
wavenumber k, the wave equation is given by Eq. (7.19) as
e + k2 E
e = 0.
∇2 E
,
!
∂ 2 Eez ∂ 2 Eez ∂ 2 Eez
+
+
+ k2 Eez = 0.
∂ x2
∂ y2
∂ z2
Eq. (8.93) reduces to
∂ 2 eez ∂ 2 eez
+ 2 + kc2 eez = 0,
∂ x2
∂y
eez (x, y) = X(x) Y (y).
1 d 2 X 1 d 2Y
+
+ kc2 = 0.
X dx2 Y dy2
(8.90)
where k is the unbounded-medium wavenumber earlier
defined as
√
(8.91)
k = ω µε .
For reasons that become clear later (in Section 8-8), the
constant kc is called the cutoff wavenumber. In view of
e and H
e now can be
Eqs. (8.89a-d), the x and y components of E
found readily—so long as we have mathematical expressions
ez . For the TE mode, Eez = 0, so all we need to know
for Eez and H
ez , and the converse is true for the TM case.
is H
(8.96)
Substituting Eq. (8.96) into Eq. (8.95) and then dividing all
terms by X(x) Y (y) leads to:
Here
kc2 = k2 − β 2 = ω 2 µε − β 2 ,
(8.95)
where kc2 is as defined by Eq. (8.90).
The form of the partial differential equation (separate,
uncoupled derivatives with respect to x and y) allows us to
assume a product solution of the form
(8.89c)
(8.89d)
(8.93)
By adopting the mathematical form given by Eq. (8.85),
namely,
Eez (x, y, z) = eez (x, y) e− jβ z ,
(8.94)
(8.89a)
(8.89b)
(8.92)
To satisfy Eq. (8.92), each of its x̂, ŷ, and ẑ components has to
be satisfied independently. Its ẑ component is given by
(8.88f)
ez
∂ Eez
∂H
,
+ ωµ
∂y
∂x
!
ez
∂ Eez
∂H
j
e
,
−β
Hx = 2 ωε
kc
∂y
∂x
!
ez
−j
∂ Eez
∂H
e
Hy = 2 ωε
.
+β
kc
∂x
∂y
j
Eey = 2
kc
8-8 TM Modes in Rectangular
Waveguide
(8.88d)
Equations (8.88a-f) incorporate the fact that differentiation
with respect to z is equivalent to multiplication by − jβ .
By manipulating these equations algebraically, we can obtain
e and H
e in terms of
expressions for the x and y components of E
their z components, namely
!
367
(8.97)
To satisfy Eq. (8.97), each of the first two terms has to equal a
constant. Hence, we define separation constants kx and ky such
that
d 2X
+ kx2 X = 0,
dx2
d 2Y
+ ky2Y = 0,
dy2
(8.98a)
(8.98b)
and
kc2 = kx2 + ky2 .
(8.99)
368
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Before proposing solutions for Eqs. (8.98a and b), we should
consider the constraints that the solutions must meet. The electric field Eez is parallel to all four walls of the waveguide. Since
E = 0 in the conducting walls, the boundary conditions require
Eez in the waveguide cavity to go to zero as x approaches 0 and
a and as y approaches 0 and b (Fig. 8-22). To satisfy these
boundary conditions, sinusoidal solutions are chosen for X(x)
and Y (y) as:
Side view
Front view
b
y
eez = X(x) Y (y) = (A cos kx x + B sin kx x)(C cos ky y + D sin ky y).
(8.100)
These forms for X(x) and Y (y) definitely satisfy the differential equations given by Eqs. (8.98a and b). The boundary
conditions for eez are
eez = 0
eez = 0
at x = 0 and a,
(8.101a)
at y = 0 and b.
(8.101b)
z
a
x
0
(a) Cross-sectional planes
y
H field
E field
b
Satisfying eez = 0 at x = 0 requires that we set A = 0, and
similarly, satisfying eez = 0 at y = 0 requires C = 0. Satisfying
eez = 0 at x = a requires
mπ
,
m = 1, 2, 3, . . .
(8.102a)
kx =
a
and similarly, satisfying eez = 0 at y = b requires
nπ
,
n = 1, 2, 3, . . .
(8.102b)
ky =
b
Consequently,
mπ x n π y sin
e− jβ z , (8.103)
Eez = eez e− jβ z = E0 sin
a
b
where E0 = BD is the amplitude of the wave in the guide.
ez = 0 for the TM mode, the transverse
Keeping in mind that H
e
e now can be obtained by applying
components of E and H
Eq. (8.103) to Eqs. (8.89a–d),
mπ x n π y − j β mπ E0 cos
sin
e− jβ z, (8.104a)
Eex = 2
kc
a
a
b
mπ x nπ y − j β nπ Eey = 2
E0 sin
cos
e− jβ z , (8.104b)
kc
b
a
b
ex = jωε nπ E0 sin mπ x cos nπ y e− jβ z , (8.104c)
H
kc2
b
a
b
mπ x n π y mπ −
j
ωε
ey =
E0 cos
sin
e− j β z .
H
kc2
a
a
b
(8.104d)
Each combination of the integers m and n represents a viable
solution—or a mode—denoted TMmn . Associated with each
mn mode are specific field distributions for the region inside
the guide. Figure 8-23 depicts the E and H field lines for the
TM11 mode across two different cross sections of the guide.
x
a
(b) Field lines for front view
0
y
H field into page
E field
b
z
0
H field out of page
(c) Field lines for side view
Figure 8-23 TM11 electric and magnetic field lines across two
cross-sectional planes.
According to Eqs. (8.103) and (8.104e), a rectangular waveguide with cross section (a × b) can support the propagation
of waves with many different, but discrete, field configurations
specified by the integers m and n. The only quantity in the
8-8 TM MODES IN RECTANGULAR WAVEGUIDE
fields’ expressions that we have yet to determine is the propagation constant β , which is contained in the exponential e− jβ z .
By combining Eqs. (8.90), (8.99), and (8.102), we obtain the
following expression for β :
r
q
mπ 2 nπ 2
2
2
−
. (8.105)
β = k − kc = ω 2 µε −
a
b
(TE and TM)
Even though the expression for β was derived for TM modes,
it is equally applicable to TE modes.
The exponential e− jβ z describes a wave traveling in the
+z direction—provided that β is real—which corresponds to
k > kc . If k < kc , β becomes imaginary: β = − jα with α
real, in which case e− jβ z = e−α z , yielding evanescent waves
characterized by amplitudes that decay rapidly with z due to
the attenuation function e−α z . Corresponding to each mode
(m, n), there is a cutoff frequency fmn at which β = 0. By
setting β = 0 in Eq. (8.105) and then solving for f , we have
r
up0 m 2 n 2
+
,
fmn =
2
a
b
(TE and TM)
(8.106)
where up0 = 1/ µε is the phase velocity of a TEM wave in
an unbounded medium with constitutive parameters ε and µ .
◮ A wave in a given mode can propagate through
the guide only if its frequency f > fmn , as only then
β = real. ◭
The mode with the lowest cutoff frequency is known as
the dominant mode. The dominant mode is TM11 among TM
modes and TE10 among TE modes. Whereas a value of zero
for m or n is allowed for TE modes, it is not for TM modes
[because if either m or n is zero, Eez in Eq. (8.103) becomes
zero and all other field components vanish as well].
By combining Eqs. (8.105) and (8.106), we can express β
in terms of fmn ,
s
1−
fmn
f
2
.
(TE and TM)
(8.107)
up0
ω
=p
. (TE and TM)
β
1 − ( fmn / f )2
Eey
Eex
βη
ZTM =
=η
=−
=
ey
ex
k
H
H
s
1−
fmn
f
2
,
(8.109)
p
where η = µ /ε is the intrinsic impedance of the dielectric
material filling the guide.
Example 8-8: Mode Properties
A TM wave propagating in a dielectric-filled waveguide of unknown permittivity has a magnetic field with the y component
given by
× sin(1.5π × 1010t − 109π z)
(mA/m).
If the guide dimensions are a = 2b = 4 cm, determine (a) the
mode numbers, (b) the relative permittivity of the material
in the guide, and (c) the phase velocity, and (d) obtain an
expression for Ex .
ey given
Solution: (a) By comparison with the expression for H
by Eq. (8.104d), we deduce that the argument of x is (mπ /a)
and the argument of y is (nπ /b). Hence,
25π =
mπ
,
4 × 10−2
100π =
nπ
,
2 × 10−2
which yield m = 1 and n = 2. Therefore, the mode is TM12 .
(b) The second sine function in the expression for Hy represents sin(ω t − β z), which means that
ω = 1.5π × 1010 (rad/s) or f = 7.5 GHz,
β = 109π (rad/m).
The phase velocity of a TE or TM wave in a waveguide is
up =
The transverse electric field consists of components Eex
and Eey given by Eqs. (8.104a and b). For a wave traveling
in the +z direction, the magnetic field associated with Eex
ey [according to the right hand rule given by Eq. (7.39a)].
is H
ex . The
Similarly, the magnetic field associated with Eey is −H
ratios, obtained by employing Eq. (8.104e), constitute the
wave impedance in the guide,
Hy = 6 cos(25π x) sin(100π y)
√
ω
β=
up0
369
By rewriting Eq. (8.105) to obtain an expression for εr = ε /ε0
in terms of the other quantities, we have
(8.108)
εr =
mπ 2 nπ 2 c2
2
β
+
,
+
ω2
a
b
370
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
where c is the speed of light. Inserting the available values, we
obtain
εr =
(3 × 108)2
(1.5π × 1010)2
"
Answer: tan(π y/a)/ tan(π x/a).
π
· (109π ) +
4 × 10−2
2
2
2π
+
2 × 10−2
2 #
= 9.
(c)
up =
which is slower than the speed of light. However, as explained
later in Section 8-10, the phase velocity in a waveguide may
exceed c, but the velocity with which energy is carried down
the guide is the group velocity ug , which is never greater than c.
(d) From Eq. (8.109),
inant TM mode in a waveguide filled with a material with
εr = 4? The waveguide dimensions are a = 2b = 5 cm.
Exercise 8-12: What is the magnitude of the phase velocity of a TE or TM mode at f = fmn ?
Answer: up = ∞ ! [See explanation in Section 8-10.]
8-9 TE Modes in Rectangular Waveguide
q
1 − ( f12 / f )2 .
Application of Eq. (8.106) yields f12 = 5.15 GHz for the TM12
mode. Using that in the expression for ZTM , in addition to
f = 7.5 GHz and
p
p
µ0 /ε0 377
= √ = 125.67 Ω,
η = µ /ε = √
εr
9
gives
ZTM = 91.3 Ω.
Hence,
Ex = ZTM Hy = 91.3 × 6 cos(25π x) sin(100π y)
× sin(1.5π × 1010t − 109π z)
(mV/m)
× sin(1.5π × 1010t − 109π z)
(V/m).
= 0.55 cos(25π x) sin(100π y)
What are the primary limitations of coaxial cables at frequencies higher than 30 GHz?
Concept Question 8-8:
Can a TE mode have a zero
magnetic field along the direction of propagation?
Concept Question 8-9:
What is the rationale for
choosing a solution for eez that involves sine and cosine
functions?
Concept Question 8-10:
Concept Question 8-11:
Exercise 8-11: What is the cutoff frequency for the dom-
Answer: For TM11 , f11 = 3.35 GHz.
ω
1.5π × 1010
=
= 1.38 × 108 m/s,
β
109π
ZTM = η
Exercise 8-10: For a square waveguide with a = b, what
is the value of the ratio Eex /Eey for the TM11 mode?
What is an evanescent wave?
In the TM case, where the wave has no magnetic field comez = 0), we started our
ponent along the z direction (i.e., H
treatment in the preceding section by obtaining a solution
for Eez , and then we used it to derive expressions for the
e and H.
e For the TE case, the same
tangential components of E
basic procedure can be applied, except we reversed the roles
ez . Such a process leads to:
of Eez and H
mπ x n π y j ω µ nπ Eex = 2
H0 cos
sin
e− jβ z, (8.110a)
kc
b
a
b
mπ x nπ y − j ω µ mπ H0 sin
cos
e− j β z ,
Eey =
2
kc
a
a
b
(8.110b)
ex = jβ mπ H0 sin mπ x cos nπ y e− jβ z , (8.110c)
H
kc2
a
a
b
ey = jβ nπ H0 cos mπ x sin nπ y e− jβ z , (8.110d)
H
kc2 b
a
b
mπ x nπ y ez = H0 cos
H
(8.110e)
cos
e− j β z ,
a
b
and of course, Eez = 0. The expressions for fmn , β , and up
given earlier by Eqs. (8.106), (8.107), and (8.108) remain
unchanged.
◮ Because not all of the fields vanish if m or n assume a
value of zero, the lowest order TE mode is TE10 if a > b or
TE01 if a < b. It is customary to assign a to be the longer
dimension, in which case the TE10 mode is the de facto
dominant mode. ◭
8-9 TE MODES IN RECTANGULAR WAVEGUIDE
371
Table 8-3 Wave properties for TE and TM modes in a rectangular waveguide with dimensions a × b that are filled with a dielectric
material with constitutive parameters ε and µ . The TEM case—shown here for reference—pertains to plane-wave propagation in
an unbounded medium.
Rectangular Waveguides
TE Modes
Eex =
Eey =
j ω µ nπ mπ x
sin nπb y e− jβ z
b H0 cos
a
kc2
− j ω µ mπ mπ x
cos nπb y e− jβ z
a H0 sin
a
kc2
Eez = 0
ex = −Eey /ZTE
H
ey = Eex /ZTE
H
ez = H0 cos mπ x cos
H
a
p
ZTE = η / 1 − ( fc / f )2
nπ y b
e− j β z
Plane Wave
TM Modes
Eex =
E0 cos maπ x sin nπb y e− jβ z
nπ
mπ x
Eey =
cos nπb y e− jβ z
b E0 sin
a
Eez = E0 sin maπ x sin nπb y e− jβ z
ex = −Eey /ZTM
H
− jβ
kc2
− jβ
kc2
mπ
a
ey = Eex /ZTM
H
ez = 0
H
p
ZTM = η 1 − ( fc / f )2
Properties Common to TE and TM Modes
r
up0 m 2 n 2
fc =
+
2
a
b
p
β = k 1 − ( fc / f )2
q
ω
up = = up0 / 1 − ( fc / f )2
β
Another difference between the TE and TM modes relates
to the expression for the wave impedance. For TE,
ZTE =
Eey
Eex
η
.
=−
=p
e
e
Hy
Hx
1 − ( fmn / f )2
TEM Mode
Eex = Ex0 e− jβ z
Eey = Ey0 e− jβ z
Eez = 0
ex = −Eey /η
H
ey = Eex /η
H
ez = 0
H
p
η = µ /ε
fc = not
applicable
√
k = ω µε
√
up0 = 1/ µε
TE02
TE11
TE10 TE01 TE20 TE21 TE30 TE31
TE12
(8.111)
A summary of the expressions for the various wave attributes
of TE and TM modes is given in Table 8-3. As a reference,
corresponding expressions for the TEM mode on a coaxial
transmission line are included as well.
fmn (GHz)
0
5
10
TM11
15
TM21
20
TM22
TM12 TM31
Figure 8-24 Cutoff frequencies for TE and TM modes in
a hollow rectangular waveguide with a = 3 cm and b = 2 cm
(Example 8-9).
Example 8-9:
Cutoff Frequencies
For a hollow rectangular waveguide with dimensions a = 3 cm
and b = 2 cm, determine the cutoff frequencies for all modes
up to 20 GHz. Over what frequency range will the guide
support the propagation of a single dominant mode?
Solution:
√ A hollow guide has µ = µ0 and ε = ε0 . Hence,
up0 = 1/ µ0 ε0 = c. Application of Eq. (8.106) gives the cutoff
frequencies shown in Fig. 8-24, which start at 5 GHz for
the TE10 mode. To avoid all other modes, the frequency of
operation should be restricted to the 5–7.5 GHz range.
372
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Gaussian pulse
High-frequency
carrier
(a)
Amplitudemodulated
waveform
encoded in it. Depending on whether or not the propagation
medium is dispersive, ug may or may not be equal to the
phase velocity up . In Section 2-1.1, we described a dispersive
transmission line as one “on which the phase velocity is not a
constant as a function of frequency,” and as a result the shape
of a pulse transmitted through it gets progressively distorted as
it moves down the line. A rectangular waveguide constitutes
a dispersive transmission line because the phase velocity of a
TE or TM mode propagating through it is a strong function of
frequency [per Eq. (8.108)], particularly at frequencies close to
the cutoff frequency fmn . As we see shortly, if f ≫ fmn , the TE
and TM modes become approximately TEM in character—not
only in terms of the directional arrangement of the electric and
magnetic fields but also in terms of the frequency dependence
of the phase velocity.
We now examine up and ug in more detail. The phase
velocity defined as the velocity of the sinusoidal pattern of the
wave is given by
ω
(8.112)
up = ,
β
while the group velocity ug is given by
(b)
Figure 8-25 The amplitude-modulated high-frequency waveform in (b) is the product of the Gaussian-shaped pulse with the
sinusoidal high-frequency carrier in (a).
8-10 Propagation Velocities
When a wave is used to carry a message through a medium
or along a transmission line, information is encoded into the
wave’s amplitude, frequency, or phase. A simple example is
shown in Fig. 8-25, where a high-frequency sinusoidal wave
of frequency f is amplitude-modulated by a low-frequency
Gaussian pulse. The waveform in Fig. 8-25(b) is the result of
multiplying the Gaussian pulse shape in Fig. 8-25(a) by the
carrier waveform.
By Fourier analysis, the waveform in Fig. 8-25(b) is equivalent to the superposition of a group of sinusoidal waves with
specific amplitudes and frequencies. Exact equivalence may
require a large (or infinite) number of frequency components,
but in practice, it is often possible to represent the modulated
waveform to a fairly high degree of fidelity with a wave group
that extends over a relatively narrow bandwidth surrounding
the high-frequency carrier f . The velocity that the envelope—
or equivalently the wave group—travels through the medium
is called the group velocity ug . As such, ug is the velocity of
the energy carried by the wave group and of the information
ug =
1
.
d β /d ω
(8.113)
Even though we will not derive Eq. (8.113) in this book, it
is nevertheless important that we understand its properties for
TE and TM modes in a metal waveguide. Using the expression
for β given by Eq. (8.107),
ug =
1
= up0
d β /d ω
q
1 − ( fmn / f )2 ,
(8.114)
where—as before—up0 is the phase velocity in an unbounded
dielectric medium. In view of Eq. (8.108) for the phase
velocity up ,
up ug = u2p0 .
(8.115)
Above cutoff ( f > fmn ), up ≥ u p0 and ug ≤ u p0 . As f → ∞, or
more precisely as ( fmn / f ) → 0, TE and TM modes approach
the TEM case for which up = ug = up0 .
A useful graphical tool for describing the propagation properties of a medium or transmission line is the ω -β diagram.
In Fig. 8-26, the straight line starting at the origin represents
the ω -β relationship for a TEM wave propagating in an
unbounded medium (or on a TEM transmission line). The
TEM line provides a reference to which the ω -β curves of
the TE/TM modes can be compared. At a given location on
the ω -β line or curve, the ratio of the value of ω to that of
8-10 PROPAGATION VELOCITIES
373
Using the identity sin θ = (e jθ − e− jθ )/2 j for any argument θ ,
we obtain
ω µπ H0
(e− jπ x/a − e jπ x/a )e− jβ z
Eey =
2kc2 a
ω (rad/s)
f21
f11
f01
f10
= E0′ (e− jβ (z+π x/β a) − e− jβ (z−π x/β a))
TM21
TE21 and
′
M 11
and T
TE 11
TE01
′′
= E0′ (e− jβ z − e− jβ z ),
where we have consolidated the quantities multiplying the two
exponential terms into the constant E0′ . The first exponential
term represents a wave with propagation constant β traveling
in the z′ direction, where
TEM
TE10
z′ = z +
β (rad/m)
Figure 8-26 ω -β diagram for TE and TM modes in a hollow
rectangular waveguide. The straight line pertains to propagation
in an unbounded medium or on a TEM transmission line.
β defines up = ω /β , whereas it is the slope d ω /d β of the
curve at that point that defines the group velocity ug . For
the TEM line, the ratio and the slope have identical values
(hence, up = ug ), and the line starts at ω = 0. In contrast, the
curve for each of the indicated TE/TM modes starts at a cutoff
frequency specific to that mode below which the waveguide
cannot support the propagation of a wave in that mode. At
frequencies close to cutoff, up and ug assume very different
values; in fact, at cutoff up = ∞ and ug = 0. On the other
end of the frequency spectrum—at frequencies much higher
than fmn —the ω -β curves of the TE/TM modes approach the
TEM line. We should note that for TE and TM modes, up
may easily exceed the speed of light, but ug will not. Since
it is ug that represents the actual transport of energy, Einstein’s
assertion that there is an upper bound on the speed of physical
phenomena is not violated.
So far, we have described the fields in the guide, but we
have yet to interpret them in terms of plane waves that zigzag
along the guide through successive reflections. To do just that,
consider the simple case of a TE10 mode. For m = 1 and n = 0,
the only nonzero component of the electric field given by
Eq. (8.110) is Eey , so
πx
ωµ π H0 sin
e− j β z .
Eey = − j 2
kc a
a
(8.117)
(8.116)
πx
,
βa
(8.118a)
and the second term represents a wave traveling in the z′′
direction with
πx
.
(8.118b)
z′′ = z −
βa
From the diagram shown in Fig. 8-27(a), it is evident that the
z′ direction is at an angle θ ′ relative to z, and the z′′ direction is
at an angle θ ′′ = −θ ′ . This means that the electric field Eey (and
e of the TE10 mode is composed
its associated magnetic field H)
of two TEM waves shown in Fig. 8-27(b) with both traveling
in the +z direction by zigzagging between the opposite walls
of the waveguide. Along the zigzag directions (z′ and z′′ ), the
phase velocity of the individual wave components is up0 , but
the phase velocity of the combination of the two waves along
z is up .
Example 8-10: Zigzag Angle
For the TE10 mode, express the zigzag angle θ ′ in terms of the
ratio ( f / f10 ), and then evaluate it at f = f10 and for f ≫ f10 .
Solution: From Fig. 8-27,
′
θ10
= tan−1
π
β10 a
,
where the subscript 10 has been added as a reminder that the
expression applies to the TE10 mode specifically. For m = 1
and n = 0, Eq. (8.106) reduces to f10 = up0 /2a. After replacing
β with the expression given by Eq. (8.107) and replacing a
with up0 /2 f10 , we obtain
"
1
#
θ ′ = tan−1 p
.
( f / f10 )2 − 1
374
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Exercise 8-13: What do the wave impedances for TE and
x
TM look like as f approaches fmn ?
Answer: At f = fmn , ZTE looks like an open circuit, and
ZTM looks like a short circuit.
z
Exercise 8-14: What are the values for (a) up , (b) ug , and
(c) the zigzag angle θ ′ at f = 2 f10 for a TE10 mode in a
hollow waveguide?
From Eq. (8.118a),
Hence,
From Eq. (8.118b),
Hence,
.
Answer: (a) up = 1.15c, (b) ug = 0.87c, (c) θ ′ = 30◦ .
.
.
8-11 Cavity Resonators
.
(a) z' and z'' propagation directions
x
z'
a
H
H
E
cos
E
z
z''
(b) TEM waves
Figure 8-27 The TE10 mode can be constructed as the sum of
two TEM waves.
At f = f10 , θ ′ = 90◦ , which means that the wave bounces back
and forth at normal incidence between the two side walls of the
waveguide, making no progress in the z direction. At the other
end of the frequency spectrum, when f ≫ f10 , θ ′ approaches
0 and the wave becomes TEM-like as it travels straight down
the guide.
For TE waves, the dominant
mode is TE10 , but for TM the dominant mode is TM11 .
Why is it not TM10 ?
Concept Question 8-12:
Why is it acceptable for up to
exceed the speed of light c, but not so for ug ?
Concept Question 8-13:
A rectangular waveguide has metal walls on four sides. When
the two remaining sides are terminated with conducting walls,
the waveguide becomes a cavity. By designing cavities to
resonate at specific frequencies, they can be used as circuit
elements in microwave oscillators, amplifiers, and bandpass
filters.
The rectangular cavity shown in Fig. 8-28(a) with dimensions (a × b × d) is connected to two coaxial cables that feed
and extract signals into and from the cavity via input and
output probes. As a bandpass filter, the function of a resonant
cavity is to block all spectral components of the input signal
except for those with frequencies that fall within a narrow
band surrounding a specific center frequency f0 , which is the
cavity’s resonant frequency. Comparison of the spectrum in
Fig. 8-28(b), which describes the range of frequencies that
might be contained in a typical input signal, with the narrow
output spectrum in Fig. 8-28(c) demonstrates the filtering
action imparted by the cavity.
In a rectangular waveguide, the fields constitute standing
waves along the x and y directions and a propagating wave
along ẑ. The terms TE and TM were defined relative to the
propagation direction; TE meant that E was entirely transverse
to ẑ, and TM meant that H had no component along ẑ.
In a cavity, there is no unique propagation direction, as no
fields propagate. Instead, standing waves exist along all three
directions. Hence, the terms TE and TM need to be modified
by defining the fields relative to one of the three rectangular
axes. For the sake of consistency, we will continue to define
the transverse direction to be any direction contained in the
plane whose normal is ẑ.
The TE mode in the rectangular waveguide consists of a
ez component is given by
single propagating wave whose H
Eq. (8.110e) as
ez = H0 cos mπ x cos nπ y e− jβ z ,
H
a
b
(8.119)
8-11 CAVITY RESONATORS
375
Hollow or dielectric-filled resonant cavity
y
b
z
d
Output
signal
x
a
ez must be zero
zero at a conducting boundary. Consequently, H
at z = 0 and z = d. To satisfy these conditions, it is necessary
that H0− = −H0 and β d = pπ , with p = 1, 2, 3, . . . . In this case,
Eq. (8.120) becomes
ez = −2 jH0 cos mπ x cos nπ y sin pπ z .
H
a
b
d
(8.121)
Given that Eez = 0 for the TE modes, all of the other compoe and H
e can be derived readily through the application
nents of E
of the relationships given by Eq. (8.89). A similar procedure
also can be used to characterize cavity modes for the TM case.
0
8-11.1 Resonant Frequency
Input signal
The consequence of the quantization condition imposed on β ,
namely β = pπ /d with p assuming only integer values, is
that for any specific set of integer values of (m, n, p) the wave
inside the cavity can exist at only a single resonant frequency,
fmnp , whose value has to satisfy Eq. (8.105). The resulting
expression for fmnp is
(a) Resonant cavity
f
f0
up
fmnp = 0
2
(b) Input spectrum
r m 2
a
+
n 2
b
+
p 2
d
.
(8.122)
For TE, the indices m and n start at 0, but p starts at 1.
The exact opposite applies to TM. By way of an example,
the resonant frequency for a TE101 mode in a hollow cavity
with dimensions a = 2 cm, b = 3 cm, and d = 4 cm is
f101 = 8.38 GHz.
∆f
f
f0
(c) Output spectrum
Figure 8-28 A resonant cavity supports a very narrow bandwidth around its resonant frequency f0 .
where the phase factor e− jβ z signifies propagation along +ẑ.
Because the cavity has conducting walls at both z = 0 and
z = d, it will contain two such waves: one with amplitude H0
traveling along +ẑ and another with amplitude H0− traveling
along −ẑ. Hence,
ez = (H0 e− jβ z + H − e jβ z ) cos mπ x cos nπ y . (8.120)
H
0
a
b
e to be
Boundary conditions require the normal component of H
8-11.2 Quality Factor
In the ideal case, if a group of frequencies is introduced into the
cavity to excite a certain TE or TM mode, only the frequency
component at exactly fmnp of that mode will survive, and all
others will attenuate. If a probe is used to couple a sample
of the resonant wave out of the cavity, the output signal will
be a monochromatic sinusoidal wave at fmnp . In practice, the
cavity exhibits a frequency response similar to that shown in
Fig. 8-28(c), which is very narrow, but it is not a perfect spike.
The bandwidth ∆ f of the cavity is defined as the frequency
range between the two frequencies
(on either side of fmnp )
√
at which the amplitude is 1/ 2 of the maximum amplitude
(at fmnp ). The normalized bandwidth, defined as ∆ f / fmnp , is
approximately equal to the reciprocal of the quality factor Q
of the cavity:
fmnp
Q≈
.
(8.123)
∆f
376
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Module 8.5 Rectangular Waveguide Upon specifying the waveguide dimensions, the frequency f , and the mode type
(TE or TM) and number, this module provides information about the wave impedance, cutoff frequency, and other wave
attributes. It also displays the electric and magnetic field distributions inside the guide.
◮ The quality factor is defined in terms of the ratio of the
energy stored in the cavity volume to the energy dissipated
in the cavity walls through conduction. ◭
For an ideal cavity with perfectly conducting walls, no energy
loss is incurred, as a result of which Q is infinite and ∆ f ≈ 0.
Metals have very high (but not infinite) conductivities, so a real
cavity with metal walls stores most of the energy coupled into
it in its volume. However, it also loses some energy to heat
conduction. A typical value for Q is on the order of 10,000,
which is much higher than can be realized with lumped RLC
circuits.
Example 8-11:
Q of a Resonant Cavity
p
where δs = 1/ π fmnp µ0 σc is the skin depth and σc is the
conductivity of the conducting walls. Design a cubic cavity
with a TE101 resonant frequency of 12.6 GHz and evaluate its
bandwidth. The cavity walls are made of copper.
Solution: For a = b = d, m = 1, n = 0,
up0 = c = 3 × 108 m/s, Eq. (8.122) simplifies to
√
3 2 × 108
(Hz).
f101 =
2a
For f101 = 12.6 GHz, this gives
a = 1.68 cm.
At f101 = 12.6 GHz, the skin depth for copper (with
σc = 5.8 × 107 S/m) is
δs =
The quality factor for a hollow resonant cavity operating in the
TE101 mode is
Q=
1
abd(a2 + d 2 )
,
3
δs [a (d + 2b) + d 3(a + 2b)]
(8.124)
p = 1, and
=
1
[π f101 µ0 σc ]1/2
1
[π × 12.6 × 109 × 4π × 10−7 × 5.8 × 107]1/2
= 5.89 × 10−7 m.
8-11 CAVITY RESONATORS
377
Upon setting a = b = d in Eq. (8.124), the expression for Q of
a cubic cavity becomes
Q=
Hence, the cavity bandwidth is
∆f ≈
1.68 × 10−2
a
=
≈ 9, 500.
3δs 3 × 5.89 × 10−7
12.6 × 109
f101
≈
≈ 1.3 MHz.
Q
9, 500
Chapter 8 Summary
Concepts
• The relations describing the reflection and transmission
behavior of a plane EM wave at the boundary between
two different media are the consequence of satisfying
the conditions of continuity of the tangential components of E and H across the boundary.
• Snell’s laws state that θi = θr and
sin θt = (n1 /n2 ) sin θi .
For media such that n2 < n1 , the incident wave is
reflected totally by the boundary when θi ≥ θc , where
θc is the critical angle given by θc = sin−1 (n2 /n1 ).
• By successive multiple reflections, light can be guided
through optical fibers. The maximum data rate of digital
pulses that can be transmitted along optical fibers is
dictated by modal dispersion.
Important Terms
• At the Brewster angle for a given polarization, the incident wave is transmitted totally across the boundary.
For nonmagnetic materials, the Brewster angle exists
for parallel polarization only.
• Any plane wave incident on a plane boundary can be
synthesized as the sum of a perpendicularly polarized
wave and a parallel polarized wave.
• Transmission-line equivalent models can be used to
characterize wave propagation, reflection by, and transmission through boundaries between different media.
• Waves can travel through a metal waveguide in the form
of transverse electric (TE) and transverse magnetic
(TM) modes. For each mode, the waveguide has a
cutoff frequency below which a wave cannot propagate.
• A cavity resonator can support standing waves at specific resonant frequencies.
Provide definitions or explain the meaning of the following terms:
ω -β diagram
acceptance angle θa
angles of incidence, reflection,
and transmission
Brewster angle θB
cladding
critical angle θc
cutoff frequency fmn
cutoff wavenumber kc
dominant mode
evanescent wave
fiber core
grazing incidence
group velocity ug
index of refraction n
modal dispersion
modes
optical fibers
parallel polarization
perpendicular polarization
phase-matching condition
plane of incidence
polarizing angle
quality factor Q
reflection coefficient Γ
reflectivity (reflectance) R
refraction angle
resonant cavity
resonant frequency
Snell’s laws
standing-wave ratio S
surface wave
total internal reflection
transmission coefficient τ
transmissivity (transmittance) T
transverse electric (TE)
polarization
transverse magnetic (TM)
polarization
unbounded-medium wavenumber
unpolarized
wavefront
378
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
Mathematical and Physical Models
Normal Incidence
Brewster Angle
r
E
η2 − η1
Γ = 0i =
η2 + η1
E0
τ=
θBk = sin
E0t
2η2
=
η2 + η1
E0i
τ = 1+Γ
√
√
εr1 − εr2
Γ= √
√
εr1 + εr2
up
sin θt
= 2 =
sin θi
up1
r
β=
(if µ1 = µ2 )
fmn
µ1 ε1
µ2 ε2
Perpendicular Polarization
Γ⊥ =
r
E⊥0
η2 cos θi − η1 cos θt
=
i
E⊥0 η2 cos θi + η1 cos θt
τ⊥ =
t
E⊥0
2η2 cos θi
=
i
η2 cos θi + η1 cos θt
E⊥0
Parallel Polarization
τk =
r
Ek0
=
η2 cos θt − η1 cos θi
η2 cos θt + η1 cos θi
t
Ek0
=
2η2 cos θi
η2 cos θt + η1 cos θi
i
Ek0
Ei
k0
PROBLEMS
Section 8-1: Reflection and Transmission at Normal Incidence
∗ 8.1 A plane wave in air with an electric field amplitude of
20 V/m is incident normally upon the surface of a lossless,
nonmagnetic medium with εr = 25. Determine:
(a) the reflection and transmission coefficients,
(b) the standing-wave ratio in the air medium, and
∗ Answer(s) available in Appendix E.
r
ω 2 µε −
mπ 2
−
a
r up
m 2
n 2
= 0
+
2
a
b
up =
Oblique Incidence
Γk =
1
= tan−1
1 + (ε1/ε2 )
r
ε2
ε1
Waveguides
Snell’s Laws
θi = θr
−1
s
nπ 2
b
up0
ω
=p
β
1 − ( fmn / f )2
up ug = u2p0
η
ZTE = p
1 − ( fmn / f )2
s
fmn 2
ZTM = η 1 −
f
Resonant Cavity
r
up0 m 2 n 2 p 2
+
+
fmnp =
2
a
b
d
fmnp
Q≈
∆f
(c) the average power densities of the incident, reflected, and
transmitted waves.
8.2 A plane wave traveling in medium 1 with εr1 = 2.25 is
normally incident upon medium 2 with εr2 = 4. Both media are
made of nonmagnetic, nonconducting materials. The electric
field of the incident wave is given by
Ei = ŷ8 cos(6π × 109t − 30π x)
(V/m).
(a) Obtain time-domain expressions for the electric and magnetic fields in each of the two media.
PROBLEMS
379
(b) Determine the average power densities of the incident,
reflected and transmitted waves.
8.3 A plane wave traveling in a medium with εr1 = 9 is
normally incident upon a second medium with εr2 = 4. Both
media are made of nonmagnetic, nonconducting materials. The
magnetic field of the incident plane wave is given by
Hi = ẑ 2 cos(2π × 109t − ky)
d
Medium 1
Medium 2
Medium 3
z
(A/m).
(a) Obtain time-domain expressions for the electric and magnetic fields in each of the two media.
∗ (b) Determine the average power densities of the incident,
reflected, and transmitted waves.
8.4 A 200 MHz, left-hand circularly polarized plane wave
with an electric field modulus of 5 V/m is normally incident
in air upon a dielectric medium with εr = 4 and occupies the
region defined by z ≥ 0.
(a) Write an expression for the electric field phasor of the
incident wave given that the field is a positive maximum
at z = 0 and t = 0.
(b) Calculate the reflection and transmission coefficients.
(c) Write expressions for the electric field phasors of the
reflected wave, the transmitted wave, and the total field
in the region z ≤ 0.
(d) Determine the percentages of the incident average power
reflected by the boundary and transmitted into the second
medium.
8.5 Repeat Problem 8.4, but replace the dielectric medium
with a poor conductor characterized by εr = 2.25, µr = 1, and
σ = 10−4 S/m.
8.6 A 50 MHz plane wave with electric field amplitude of
50 V/m is normally incident in air onto a semi-infinite, perfect
dielectric medium with εr = 36. Determine:
∗(a) Γ,
(b) the average power densities of the incident and reflected
waves, and
(c) the distance in the air medium from the boundary to the
nearest minimum of the electric field intensity, |E|.
∗ 8.7 What is the maximum amplitude of the total electric field
in the air medium of Problem 8.6, and at what nearest distance
from the boundary does it occur?
8.8 Repeat Problem 8.6, but replace the dielectric medium
with a conductor with εr = 1, µr = 1, and σ = 2.78 × 10−3
S/m.
∗ 8.9 The three regions shown in Fig. P8.9 contain perfect
dielectrics. For a wave in medium 1, incident normally upon
εr2
ε r1
z = −d
εr3
z=0
Figure P8.9 Dielectric layers for Problems 8.9 to 8.11.
the boundary at z = −d, what combination of εr2 and d
produces no reflection? Express your answers in terms of εr1 ,
εr3 and the oscillation frequency of the wave, f .
8.10 For the configuration shown in Fig. P8.9, use
transmission-line equations (or the Smith chart) to calculate
the input impedance at z = −d for εr1 = 1, εr2 = 9, εr3 = 4,
d = 1.2 m, and f = 50 MHz. Also determine the fraction of
the incident average power density reflected by the structure.
Assume all media are lossless and nonmagnetic.
∗ 8.11 Repeat Problem 8.10, but interchange ε and ε .
r1
r3
8.12 Orange light of wavelength 0.61 µ m in air enters a
block of glass with εr = 1.44. What color would it appear to
a sensor embedded in the glass? The wavelength ranges of
colors are violet (0.39 to 0.45 µ m), blue (0.45 to 0.49 µ m),
green (0.49 to 0.58 µ m), yellow (0.58 to 0.60 µ m), orange
(0.60 to 0.62 µ m), and red (0.62 to 0.78 µ m).
∗ 8.13 A plane wave of unknown frequency is normally inci-
dent in air upon the surface of a perfect conductor. Using an
electric-field meter, it was determined that the total electric
field in the air medium is always zero when measured at a
distance of 2 m from the conductor surface. Moreover, no such
nulls were observed at distances closer to the conductor. What
is the frequency of the incident wave?
8.14 Consider a thin film of soap in air under illumination by
yellow light with λ = 0.6 µ m in vacuum. The film is treated
as a planar dielectric slab with εr = 1.72 and is surrounded on
both sides by air. What film thickness would produce strong
reflection of the yellow light at normal incidence?
∗ 8.15 A 5 MHz plane wave with electric field amplitude of
10 (V/m) is normally incident in air onto the plane surface
of a semi-infinite conducting material with εr = 4, µr = 1,
380
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
and σ = 100 (S/m). Determine the average power dissipated
(lost) per unit cross-sectional area in a 2 mm penetration of the
conducting medium.
8.16 A 0.5 MHz antenna carried by an airplane flying over
the ocean surface generates a wave that approaches the water
surface in the form of a normally incident plane wave with
an electric-field amplitude of 3,000 (V/m). Seawater is characterized by εr = 72, µr = 1, and σ = 4 (S/m). The plane is
trying to communicate a message to a submarine submerged at
a depth d below the water surface. If the submarine’s receiver
requires a minimum signal amplitude of 0.01 (µ V/m), what is
the maximum depth d where successful communication is still
possible?
Angular dispersion
60°
Red
50°
Green
Vio
let
Figure P8.18 Prism of Problem 8.18.
Sections 8-2 and 8-3: Snell’s Laws and Fiber Optics
∗ 8.17 A light ray is incident on a prism in air at an angle θ ,
as shown in Fig. P8.17. The ray is refracted at the first surface
and again at the second surface. In terms of the apex angle φ of
the prism and its index of refraction n, determine the smallest
value of θ for which the ray will emerge from the other side.
Find this minimum θ for n = 1.4 and φ = 60◦.
45°
90°
Su
φ
90°
2
ce
rfa
θ
Su
rfa
ce
1
45°
45°
45°
Figure P8.19 Periscope prisms of Problem 8.19.
n
Figure P8.17 Prism of Problem 8.17.
8.18 For some types of glass, the index of refraction varies
with wavelength. A prism is made of a material with
n = 1.71 −
4
λ0
30
(λ0 in µ m),
where λ0 is the wavelength in vacuum, and is used to disperse
white light, as shown in Fig. P8.18. The white light is incident
at an angle of 50◦ , the wavelength λ0 of red light is 0.7 µ m, and
that of violet light is 0.4 µ m. Determine the angular dispersion
in degrees.
∗ 8.19 The two prisms in Fig. P8.19 are made of glass with
n = 1.5. What fraction of the power density carried by the ray
incident upon the top prism emerges from the bottom prism?
Neglect multiple internal reflections.
8.20 A parallel-polarized plane wave is incident from air at
an angle θi = 30◦ onto a pair of dielectric layers, as shown in
Fig. P8.20.
(a) Determine the angles of transmission θ2 , θ3 , and θ4 .
(b) Determine the lateral distance d.
8.21 A light ray incident at 45◦ passes through two dielectric
materials with the indices of refraction and thicknesses given
in Fig. P8.21. If the ray strikes the surface of the first dielectric
at a height of 2 cm, at what height will it strike the screen?
∗ 8.22 Figure P8.22 depicts a beaker containing a block of
glass on the bottom and water over it. The glass block contains
a small air bubble at an unknown depth below the water
surface. When viewed from above at an angle of 60◦ , the air
bubble appears at a depth of 6.81 cm. What is the true depth of
the air bubble?
PROBLEMS
381
θi
60°
Air
μr = 1
εr = 6.25
5 cm
θ2
μr = 1
εr = 2.25
θ3
5 cm
θ4
10 cm
6.81 cm
Water
n = 1.33
Apparent position
of air bubble
Air
d
Air bubble
Glass
n = 1.6
Figure P8.20 Problem P8.20.
Figure P8.22 Apparent position of the air bubble in Problem 8.22.
n = 1 n = 1.5 n = 1.3
n=1
Oil drop
noil
Screen
2 cm
θ
45°
3 cm
4 cm
5 cm
nglass
Figure P8.21 Light incident on a screen through a multilayered dielectric (Problem 8.21).
8.23 A glass semicylinder with n = 1.5 is positioned such
that its flat face is horizontal, as shown in Fig. P8.23, and its
horizontal surface supports a drop of oil, as also shown. When
light is directed radially toward the oil, total internal reflection
occurs if θ exceeds 53◦ . What is the index of refraction of the
oil?
∗ 8.24 A penny lies at the bottom of a water fountain at a depth
of 30 cm. Determine the diameter of a piece of paper that when
placed to float on the surface of the water directly above the
penny, would totally obscure the penny from view. Treat the
penny as a point and assume that n = 1.33 for water.
8.25 Suppose that the optical fiber of Example 8-5 is submerged in water (with n = 1.33) instead of air. Determine θa
and fp in that case.
Figure P8.23 Oil drop on the flat surface of a glass semicylinder (Problem 8.23).
∗ 8.26 Equation (8.45) was derived for the case where the light
incident upon the sending end of the optical fiber extends over
the entire acceptance cone shown in Fig. 8-12(b). Suppose
the incident light is constrained to a narrower range extending
between normal incidence and θ ′ , where θ ′ < θa .
(a) Obtain an expression for the maximum data rate fp in
terms of θ ′ .
(b) Evaluate fp for the fiber of Example 8-5 when θ ′ = 5◦ .
Sections 8-4 and 8-5: Reflection and Transmission at Oblique
Incidence
8.27 A 2-km long optical fiber uses a fiber core with nf = 1.6
and a cladding with nc = 1.57. Compute the maximum data
382
CHAPTER 8 WAVE REFLECTION AND TRANSMISSION
rate and confirm your result using Module 8.2. The operating
frequency is 100 THz.
8.28 A 3-km long optical fiber uses a fiber core with
nf = 1.51 and a cladding with nc = 1.48. Compute the maximum data rate and confirm your result using Module 8.2. The
operating frequency is 300 THz.
8.29 A plane wave in air with
e i = ŷ 20e− j(3x+4z)
E
waves. Determine the fraction of the incident power reflected
by the planar surface of a piece of glass with n = 1.5 when
illuminated by natural light at 70◦ .
∗ 8.33 A parallel-polarized plane wave is incident from air
onto a dielectric medium with εr = 9 at the Brewster angle.
What is the refraction angle?
8.34 Repeat Problem 8.33 using Module 8.3.
(V/m)
is incident upon the planar surface of a dielectric material with
εr = 4, and it occupies the half-space z ≥ 0. Determine:
(a) the polarization of the incident wave,
∗ (b) the angle of incidence,
(c) the time-domain expressions for the reflected electric and
magnetic fields,
(d) the time-domain expressions for the transmitted electric
and magnetic fields, and
(e) the average power density carried by the wave in the
dielectric medium.
8.35 A perpendicularly polarized wave in air is obliquely
incident upon a planar glass–air interface at an incidence angle
of 30◦ . The wave frequency is 600 THz (1 THz = 1012 Hz),
which corresponds to green light, and the index of refraction
of the glass is 1.6. If the electric field amplitude of the incident
wave is 50 V/m, determine:
(a) the reflection and transmission coefficients, and
(b) the instantaneous expressions for E and H in the glass
medium.
8.36 Repeat part (a) of Problem 8.35 using Module 8.3.
8.37 Show that the reflection coefficient Γ⊥ can be written in
the following form:
8.30 Repeat Problem 8.29 for a wave in air with
e i = ŷ 2 × 10−2e− j(8x+6z)
H
(A/m)
incident upon the planar boundary of a dielectric medium
(z ≥ 0) with εr = 9.
8.31 A plane wave in air with
e i = (x̂ 9 − ŷ4 − ẑ6)e− j(2x+3z)
E
(V/m)
is incident upon the planar surface of a dielectric material, with
εr = 2.25, and it occupies the half-space z ≥ 0. Determine:
∗ (a) the incidence angle θ ,
i
(b) the frequency of the wave,
e r of the reflected wave,
(c) the field E
e t of the wave transmitted into the dielectric
(d) the field E
medium, and
(e) the average power density carried by the wave into the
dielectric medium.
8.32 Natural light is randomly polarized, which means that,
on average, half the light energy is polarized along any given
direction (in the plane orthogonal to the direction of propagation) and the other half of the energy is polarized along the
direction orthogonal to the first polarization direction. Hence,
when treating natural light incident upon a planar boundary,
we can consider half of its energy to be in the form of parallelpolarized waves and the other half as perpendicularly polarized
Γ⊥ =
sin(θt − θi )
sin(θt + θi )
8.38 Show that for nonmagnetic media, the reflection coefficient Γk can be written in the following form:
Γk =
tan(θt − θi )
.
tan(θt + θi )
∗ 8.39 A parallel-polarized beam of light with an electric
field amplitude of 10 (V/m) is incident in air on polystyrene
with µr = 1 and εr = 2.6. If the incidence angle at the air–
polystyrene planar boundary is 50◦ , determine:
(a) the reflectivity and transmissivity, and
(b) the power carried by the incident, reflected, and transmitted beams if the spot on the boundary illuminated by the
incident beam is 1 m2 in area.
8.40 Repeat part (a) of Problem 8.39 using Module 8.3.
8.41 A 50-MHz right-hand circularly polarized plane wave
with an electric field modulus of 30 V/m is normally incident
in air upon a dielectric medium with εr = 9 and occupying the
region defined by z ≥ 0.
(a) Write an expression for the electric field phasor of the
incident wave given that the field is a positive maximum
at z = 0 and t = 0.
PROBLEMS
383
(b) Calculate the reflection and transmission coefficients.
(c) Write expressions for the electric field phasors of the
reflected wave, the transmitted wave, and the total field
in the region z ≤ 0.
(d) Determine the percentages of the incident average power
reflected by the boundary and transmitted into the second
medium.
8.42 Consider a flat 5-mm thick slab of glass with εr = 2.56.
∗(a) If a beam of green light (λ = 0.52 µ m) is normally
0
incident upon one of the sides of the slab, what percentage
of the incident power is reflected back by the glass?
(b) To eliminate reflections, it is desired to add a thin layer
of antireflection coating material on each side of the
glass. If you are at liberty to specify the thickness of the
antireflection material as well as its relative permittivity,
what would these specifications be?
8.43 A perpendicular polarization wave in air is incident
upon the surface of a perfect conductor at an incidence angle
of 30◦ . The frequency of the wave is 2 GHz. Use Module 8.1
to determine the location of the nearest electric field maximum
from the conductor surface.
8.44 A 2 GHz plane wave in air is incident upon a conducting
medium with εr = 4 and σ = 2 S/m at an incidence angle
of 60◦. The wave is parallel-polarized. Use Module 8.4 to
compute:
(a) the reflectivity and transmissivity, and
(b) the effective refraction angle.
8.45 A 2 GHz plane wave in air is incident upon a conducting
medium with εr = 4 and σ = 2 S/m at an incidence angle
of 60◦ . The wave is perpendicularly polarized. Use Module
8.4 to compute:
(a) the reflectivity and transmissivity, and
(b) the effective refraction angle.
Sections 8-6 to 8-11: Waveguides and Resonators
determine:
(a) the mode number,
(b) εr of the material in the guide,
(c) the cutoff frequency, and
(d) the expression for Hy .
∗ 8.49 A waveguide filled with a material whose ε = 2.25 has
r
dimensions a = 2 cm and b = 1.4 cm. If the guide is to transmit
10.5 GHz signals, what possible modes can be used for the
transmission?
8.50 For a rectangular waveguide operating in the TE10
mode, obtain expressions for the surface charge density ρes
and surface current density e
Js on each of the four walls of the
guide.
∗ 8.51 A waveguide with dimensions a = 1 cm and b = 0.7 cm
is to be used at 20 GHz. Determine the wave impedance for the
dominant mode when
(a) the guide is empty, and
(b) the guide is filled with polyethylene (whose εr = 2.25).
8.52 A narrow rectangular pulse superimposed on a carrier
with a frequency of 9.5 GHz was used to excite all possible
modes in a hollow guide with a = 3 cm and b = 2.0 cm. If
the guide is 100 m in length, how long will it take each of the
excited modes to arrive at the receiving end?
∗ 8.53 If the zigzag angle θ ′ is 25◦ for the TE mode, what
10
would it be for the TE20 mode?
8.54 Measurement of the TE101 frequency response of an airfilled cubic cavity revealed that its Q is 4802. If its volume
is 64 mm3 , what material are its sides made of? [Hint: See
Appendix B.]
8.55 A hollow cavity made of aluminum has dimensions
a = 4 cm and d = 3 cm. Calculate Q of the TE101 mode for
∗ (a) b = 2 cm, and
8.46 Derive Eq. (8.89b).
(b) b = 3 cm.
∗ 8.47 A hollow, rectangular waveguide is to be used to
8.56 A hollow rectangular waveguide with dimensions
a = 2 cm and b = 1 cm is used to propagate signals at 20 GHz.
Use Module 8.5 to determine:
(a) the cutoff frequency of TE10 mode, and
(b) all possible transmission modes.
8.48 A TE wave propagating in a dielectric-filled waveguide of unknown permittivity has dimensions a = 5 cm and
b = 3 cm. If the x component of its electric field is given by
8.57 A hollow rectangular waveguide with dimensions
a = 4 cm and b = 2 cm is used to propagate signals at 12 GHz.
Use Module 8.5 to determine:
(a) the cutoff frequency of TE10 mode, and
(b) all possible transmission modes.
transmit signals at a carrier frequency of 6 GHz. Choose its
dimensions so that the cutoff frequency of the dominant TE
mode is lower than the carrier by 25% and that of the next
mode is at least 25% higher than the carrier.
Ex = −36 cos(40π x) sin(100π y)
· sin(2.4π × 1010t − 52.9π z),
(V/m)
Chapter
9
Radiation and Antennas
Chapter Contents
9-1
9-2
9-3
9-4
9-5
9-6
TB17
9-7
9-8
9-9
9-10
9-11
Objectives
Upon learning the material presented in this chapter, you
should be able to:
Overview, 395
The Hertzian Dipole, 387
Antenna Radiation Characteristics, 390
Half-Wave Dipole Antenna, 397
Dipole of Arbitrary Length, 400
Effective Area of a Receiving Antenna, 401
Friis Transmission Formula, 403
Health Risks of EM Fields, 405
Radiation by Large-Aperture Antennas, 408
Rectangular Aperture with Uniform Aperture
Distribution, 410
Antenna Arrays, 412
N-Element Array with Uniform Phase
Distribution, 419
Electronic Scanning of Arrays, 421
Chapter 9 Summary, 426
Problems, 428
1. Calculate the electric and magnetic fields of waves radiated by a dipole antenna.
2. Characterize the radiation of an antenna in terms of its
radiation pattern, directivity, beamwidth, and radiation
resistance.
3. Apply the Friis transmission formula to a free-space
communication system.
4. Calculate the electric and magnetic fields of waves radiated by aperture antennas.
5. Calculate the radiation pattern of multi-element antenna
arrays.
384
Overview
Electric field lines
of radiated wave
An antenna is a transducer that converts a guided wave propagating on a transmission line into an electromagnetic wave
propagating in an unbounded medium (usually free space), or
vice versa. Figure 9-1 shows how a wave is launched by a
hornlike antenna with the horn acting as the transition segment
between the waveguide and free space.
Antennas are made in various shapes and sizes (Fig. 9-2)
and are used in radio and television broadcasting and reception, radio-wave communication systems, cellular telephones,
radar systems, anticollision automobile sensors, and many
other applications. The radiation and impedance properties
of an antenna are governed by its shape, size, and material
properties. The dimensions of an antenna are usually measured
in units of λ of the wave it is launching or receiving; a 1-m long
dipole antenna operating at a wavelength λ = 2 m exhibits the
same properties as a 1-cm long dipole operating at λ = 2 cm.
Hence, in most of our discussions in this chapter, we refer to
antenna dimensions in wavelength units.
Antenna
Transmission line
Generator Guided EM wave
Transition
region
Wave launched
into free space
(a) Transmission mode
Antenna
Transmission line
Rec
Detector Guided EM wave
or receiver
Reciprocity
The directional function characterizing the relative distribution
of power radiated by an antenna is known as the antenna radiation pattern (or simply the antenna pattern). An isotropic
antenna is a hypothetical antenna that radiates equally in all
directions, and it is often used as a reference radiator when
describing the radiation properties of real antennas.
Transition
region
Incident
wave
(b) Reception mode
Figure 9-1 Antenna as a transducer between a guided electromagnetic wave and a free-space wave, for both transmission and
reception.
◮ Most antennas are reciprocal devices, exhibiting the
same radiation pattern for transmission as for reception. ◭
is used in the transmission mode, which is also called the
antenna polarization.
Reciprocity means that, if in the transmission mode a given
antenna transmits in direction A 100 times the power it transmits in direction B, then when used in the reception mode
it is 100 times more sensitive to electromagnetic radiation
incident from direction A than from B. All of the antennas
shown in Fig. 9-2 obey the reciprocity law, but not all antennas
are reciprocal devices. Reciprocity may not hold for some
solid-state antennas composed of nonlinear semiconductors or
ferrite materials. Such nonreciprocal antennas are beyond the
scope of this chapter; hence, reciprocity is assumed throughout. The reciprocity property is very convenient because it
allows us to compute the radiation pattern of an antenna in
the transmission mode—even when the antenna is intended to
operate as a receiver.
To fully characterize an antenna, one needs to study its
radiation properties and impedance. The radiation properties
include its directional radiation pattern and the associated
polarization state of the radiated wave when the antenna
◮ Being a reciprocal device, an antenna, when operating
in the receiving mode, can extract from an incident wave
only that component of the wave whose electric field
matches the antenna polarization state. ◭
The second aspect, the antenna impedance, pertains to the
transfer of power from a generator to the antenna when the
antenna is used as a transmitter and, conversely, the transfer of
power from the antenna to a load when the antenna is used as a
receiver, as will be discussed later in Section 9-5. It should
be noted that throughout our discussions in this chapter it
will be assumed that the antenna is properly matched to the
transmission line connected to its terminals, thereby avoiding
reflections and their associated problems.
385
386
CHAPTER 9 RADIATION AND ANTENNAS
Circular
plate
reflector
(a) Thin dipole
(b) Biconical dipole
(c) Loop
(d) Helix
Radiating strip
(e) Log-periodic
Phase shifters
Coaxial feed Dielectric
substrate
Feed
point
Ground metal
plane
(f) Parabolic dish
reflector
(g) Horn
(h) Microstrip
(i) Antenna array
Figure 9-2 Various types of antennas.
Radiation Sources
Radiation sources fall into two categories: currents and aperture fields. The dipole and loop antennas (Fig. 9-2(a) and (c))
are examples of current sources; the time-varying currents
flowing in the conducting wires give rise to the radiated electromagnetic fields. A horn antenna (Fig. 9-2(g)) is an example
of the second group because the electric and magnetic fields
across the horn’s aperture serve as the sources of the radiated
fields. The aperture fields are themselves induced by timevarying currents on the surfaces of the horn’s walls. Therefore,
all radiation ultimately is due to time-varying currents. The
choice of currents or apertures as the sources is merely a
computational convenience arising from the structure of the
antenna. We will examine the radiation processes associated
with both types of sources.
Far-Field Region
The wave radiated by a point source is spherical in nature, with
the wavefront expanding outward at a rate equal to the phase
velocity up (or the velocity of light c if the medium is free
space). If R, which is the distance between the transmitting
antenna and the receiving antenna, is sufficiently large enough
for the wavefront across the receiving aperture to be considered planar (Fig. 9-3), then the receiving aperture is said to be
Source
R
Transmitting
antenna
Receiving
antenna
Spherical wave
Plane-wave
approximation
Figure 9-3 Far-field plane-wave approximation.
in the far-field (or far-zone) region of the transmitting point
source. This region is of particular significance because—
for most applications—the location of the observation point
is indeed in the far-field region of the antenna. The far-field
plane-wave approximation allows the use of certain mathematical approximations that simplify the computation of the
radiated field and, conversely, provide convenient techniques
for synthesizing the appropriate antenna structure that would
give rise to the desired far-field antenna pattern.
9-1 THE HERTZIAN DIPOLE
387
Antenna Arrays
z
When multiple antennas operate together, the combination is
called an antenna array (Fig. 9-2(i)), and the array as a whole
behaves as if it were a single antenna. By controlling the
magnitude and phase of the signal feeding each antenna, it
is possible to shape the radiation pattern of the array and to
electronically steer the direction of the beam electronically.
These topics are treated in Sections 9-9 to 9-11.
Q = (R, θ, φ)
θ
R'
R
i(t)
y
l
i(t)
9-1 The Hertzian Dipole
x
By regarding a linear antenna as consisting of a large number
of infinitesimally short conducting elements where each is
so short that current may be considered uniform over its
length, the field of the entire antenna may be obtained by
integrating the fields from all these differential antennas with
the proper magnitudes and phases taken into account. We shall
first examine the radiation properties of such a differential
antenna, known as a Hertzian dipole, and in Section 9-3, we
will extend the results to compute the fields radiated by a halfwave dipole, which is commonly used as a standard antenna
for many applications.
◮ A Hertzian dipole is a thin, linear conductor whose
length l is very short compared with the wavelength λ ;
l should not exceed λ /50. ◭
In this case, the wire is oriented along the z direction in
Fig. 9-4 and carries a sinusoidally varying current given by
i(t) = I0 cos ω t = Re[I0 e jω t ]
(A),
Z
υ′
′
e
Je− jkR
dυ ′,
R′
Figure 9-4 Short dipole placed at the origin of a spherical
coordinate system.
e
J = ẑ(I0 /s), where s is the cross-sectional area of the dipole
wire. Also, d υ ′ = s dz, and the limits of integration are from
z = −l/2 to z = l/2. In Fig. 9-4, the distance R′ between the
observation point and a given point along the dipole is not the
same as the distance to its center, R, but because we are dealing
with a very short dipole, we can set R′ ≈ R. Hence,
− jkR Z
e
µ0 e− jkR l/2
µ0
e
,
(9.3)
ẑI0 dz = ẑ I0 l
A=
4π R
4π
R
−l/2
◮ The function (e− jkR/R) is called the spherical propagation factor. It accounts for the 1/R decay of the magnitude
with distance as well as the phase change represented by
e− jkR . ◭
(9.1)
where I0 is the current amplitude. From Eq. (9.1), the phasor
current Ie = I0 . Even though the current has to go to zero at the
two ends of the dipole, we shall treat it as constant across its
entire length.
The customary approach for finding the electric and magnetic fields at a point Q in space (Fig. 9-4) due to radiation
by a current source is through the retarded vector potential A.
e
From Eq. (6.84), the phasor retarded vector potential A(R)
at
′
a distance vector R from a volume υ containing a phasor
current distribution e
J is given by
µ0
e
A(R)
=
4π
φ
(9.2)
where µ0 is the magnetic permeability of free space (because
the observation point is in air) and k = ω /c = 2π /λ is the
wavenumber. For the dipole, the current density is simply
e is the same as that of the current (z direcThe direction of A
tion).
Because our objective is to characterize the directional
character of the radiated power at a fixed distance R from
the antenna, antenna pattern plots are presented in a spherical
coordinate system (Fig. 9-5). Its variables, R, θ , and φ , are
called the range, zenith angle, and azimuth angle, respece in terms of its spherical
tively. To that end, we need to write A
coordinate components, which is realized with the help of
Eq. (3.65c) by expressing ẑ in terms of spherical coordinates:
ẑ = R̂ cos θ − θ̂θ sin θ .
(9.4)
Upon substituting Eq. (9.4) into Eq. (9.3), we obtain
− jkR e = (R̂ cos θ − θ̂θ sin θ ) µ0 I0 l e
eR + θ̂θ A
eθ + φ̂φ A
eφ ,
= R̂A
A
4π
R
(9.5)
388
CHAPTER 9 RADIATION AND ANTENNAS
θ = 0°
z
S
Direction (θ, φ)
p
where η0 = µ0 /ε0 ≈ 120π (Ω) is the intrinsic impedance of
eR , H
eθ , and Eeφ ) are
free space. The remaining components (H
everywhere zero. Figure 9-6 depicts the electric field lines of
the wave radiated by the short dipole.
θ
Radiation
source
θ = 90°
φ = 270°
9-1.1
R
θ = 90° y
φ = 90°
φ
x
θ = 90°
φ = 0°
As was stated earlier, in most antenna applications, we are
primarily interested in the radiation pattern of the antenna at
great distances from the source. For the electric dipole, this
corresponds to distances R so that R ≫ λ or, equivalently,
kR = 2π R/λ ≫ 1. This condition allows us to neglect the
terms varying as 1/(kR)2 and 1/(kR)3 in Eqs. (9.8a) to (9.8c)
in favor of the terms varying as 1/kR, which yields the far-field
expressions
θ = 180°
jI0 lkη0
Eeθ =
4π
Figure 9-5 Spherical coordinate system.
e
eφ = Eθ
H
η0
with
e− jkR
,
R
− jkR e
µ0 I0 l
e
Aθ = −
sin θ
,
4π
R
eR = µ0 I0 l cos θ
A
4π
(9.6a)
(9.6b)
eφ = 0.
A
e known, the next step is
With the spherical components of A
straightforward; we simply apply the free-space relationships
given by Eqs. (6.85) and (6.86) as
and
Far-Field Approximation
e
e = 1 ∇×
×A
H
µ0
e=
E
1
e
×H
∇×
jωε0
to obtain the expressions
1
I0 lk2 − jkR j
e
Hφ =
+
sin θ ,
e
4π
kR (kR)2
1
j
2I0 lk2
− jkR
e
−
η0 e
cos θ ,
ER =
4π
(kR)2 (kR)3
I0 lk2
j
j
1
Eeθ =
sin θ ,
−
η0 e− jkR
+
4π
kR (kR)2 (kR)3
(9.7a)
(9.7b)
(9.8a)
e− jkR
R
sin θ
(V/m),
(A/m),
(9.9a)
(9.9b)
and EeR is negligible. At the observation point Q (Fig. 9-4), the
wave now appears similar to a uniform plane wave with its
electric and magnetic fields in phase, related by the intrinsic
impedance of the medium η0 , and their directions orthogonal
to each other and to the direction of propagation (R̂). Both
fields are proportional to sin θ and independent of φ (which
is expected from symmetry considerations).
9-1.2
Power Density
e and H,
e the time-average Poynting vector of the
Given E
radiated wave, which is also called the power density, can be
obtained by applying Eq. (7.100); that is,
e×H
e∗
Sav = 21 Re E×
(W/m2 ).
(9.10)
For the short dipole, use of Eqs. (9.9a) and (9.9b) yields
(9.8b)
Sav = R̂ S(R, θ ),
(9.8c)
(9.11)
9-1 THE HERTZIAN DIPOLE
389
Dipole
axis
λ
2λ
3λ
4λ
Broadside
direction
Figure 9-6 Electric field lines surrounding an oscillating dipole at a given instant.
with
S(R, θ ) =
η0 k2 I02 l 2
32π 2R2
= S0 sin2 θ
sin2 θ
(W/m2 ).
(9.12)
The directional pattern of any antenna is described in terms of
the normalized radiation intensity F(θ , φ ) and is defined as
the ratio of the power density S(R, θ , φ ) at a specified range R
to Smax , which is the maximum value of S(R, θ , φ ) at the same
range. Thus,
F(θ , φ ) =
S(R, θ , φ )
.
Smax
(dimensionless)
(9.13)
For the Hertzian dipole, the sin2 θ dependence in Eq. (9.12)
indicates that the radiation is maximum in the broadside
direction (θ = π /2), corresponding to the azimuth plane, and
is given by
η0 k2 I02 l 2 15π I02 l 2
(W/m2 ), (9.14)
Smax = S0 =
=
32π 2 R2
R2
λ
where use was made of the relations k = 2π /λ and η0 ≈ 120π .
We observe that Smax is directly proportional to I02 and l 2 (with
l measured in wavelengths) and that it decreases with distance
as 1/R2 .
From the definition of the normalized radiation intensity
given by Eq. (9.13), it follows that
F(θ , φ ) = F(θ ) = sin2 θ .
(9.15)
Plots of F(θ ) are shown in Fig. 9-7 in both the elevation plane
(the θ plane) and the azimuth plane (φ plane).
390
CHAPTER 9 RADIATION AND ANTENNAS
θ1 = 45°
F(θ)
θ
Dipole
1
θ = 90° (broadside)
β = 90° 1
0
Outline the basic steps used to
relate the current in a wire to the radiated power density.
Concept Question 9-4:
z
0.5
θ2 = 135°
Exercise 9-1: A 1-m long dipole is excited by a 5 MHz
current with an amplitude of 5 A. At a distance of 2 km,
what is the power density radiated by the antenna along
its broadside direction?
Answer: S0 = 8.2 × 10−8 W/m2 . (See
EM
.)
(a) Elevation pattern
9-2 Antenna Radiation Characteristics
y
1
F(φ)
φ
1
0
x
Dipole
An antenna pattern describes the far-field directional properties of an antenna when measured at a fixed distance from the
antenna. In general, the antenna pattern is a three-dimensional
plot that displays the strength of the radiated field or power
density as a function of direction—with direction being specified by the zenith angle θ and the azimuth angle φ .
◮ By virtue of reciprocity, a receiving antenna has the
same directional antenna pattern as the pattern that it
exhibits when operated in the transmission mode. ◭
(b) Azimuth pattern
Figure 9-7 Radiation patterns of a short dipole.
Consider a transmitting antenna placed at the origin of the
observation sphere shown in Fig. 9-8. The differential power
radiated by the antenna through an elemental area dA is
dPrad = Sav · dA = Sav · R̂ dA = S dA
◮ No energy is radiated by the short dipole along the
direction of the dipole axis, and maximum radiation
(F = 1) occurs in the broadside direction (θ = 90◦ ). Since
F(θ ) is independent of φ , the pattern is doughnut-shaped
in θ –φ space. ◭
z
(W),
(9.16)
R sin θ dφ
R dθ
dA = R2 sin θ dθ dφ
= R2 dΩ
Elevation
plane
θ R
What does it mean to say that
most antennas are reciprocal devices?
Concept Question 9-1:
y
φ
What is the radiated wave like
in the far-field region of the antenna?
Concept Question 9-2:
R dφ
Azimuth plane
In a Hertzian dipole, what is
the underlying assumption about the current flowing
through the wire?
Concept Question 9-3:
x
Figure 9-8 Definition of solid angle dΩ = sin θ d θ d φ .
9-2 ANTENNA RADIATION CHARACTERISTICS
391
Module 9.1 Hertzian Dipole (l ≪ λ ) For a short dipole oriented along the z axis, this module displays the field
distributions for E and H in both the horizontal and vertical planes. It can also animate the radiation process and current
flow through the dipole.
where S is the radial component of the time-average Poynting
vector Sav . In the far-field region of any antenna, Sav is always
in the radial direction. In a spherical coordinate system,
dA = R2 sin θ d θ d φ ,
(9.17)
Note that, whereas a planar angle is measured in radians and
the angular measure of a complete circle is 2π (rad), a solid
angle is measured in steradians (sr), and the angular measure
for a spherical surface is Ω = (4π R2)/R2 = 4π (sr). The solid
angle of a hemisphere is 2π (sr).
Using the relation dA = R2 dΩ, dPrad can be rewritten as
and the solid angle dΩ associated with dA, which is defined as
the subtended area divided by R2 , is given by
dPrad = R2 S(R, θ , φ ) dΩ.
dΩ =
dA
= sin θ d θ d φ
R2
(sr).
(9.19)
(9.18)
The total power radiated by an antenna through a spherical surface at a fixed distance R is obtained by integrating
392
CHAPTER 9 RADIATION AND ANTENNAS
Eq. (9.19) over that surface:
φ =0 θ =0
= R2 Smax
= R2 Smax
0
S(R, θ , φ ) sin θ d θ d φ
Z 2π Z π
φ =0 θ =0
ZZ
4π
F(θ , φ ) sin θ d θ d φ
F(θ , φ ) dΩ
(W),
(9.20)
where F(θ , φ ) is the normalized radiation intensity defined by
Eq. (9.13). The 4π symbol under the integral sign is used as an
abbreviation for the indicated limits on θ and φ . Formally, Prad
is called the total radiated power.
−20
−25
−30
88
89
2
1
90
0
91
s)
ee
r
eg
(d
As an example, the antenna pattern shown in Fig. 9-10(a) is
plotted on a decibel scale in polar coordinates with its intensity
as the radial variable. This format permits a convenient visual
interpretation of the directional distribution of the radiation
lobes.
−15
θ
le
F (dB) = 10 log F.
−10
g
an
Each specific combination of the zenith angle θ and the
azimuth angle φ denotes a specific direction in the spherical
coordinate system of Fig. 9-8. The normalized radiation intensity F(θ , φ ) characterizes the directional pattern of the energy
radiated by an antenna, and a plot of F(θ , φ ) as a function
of both θ and φ constitutes a three-dimensional pattern—an
example of which is shown in Fig. 9-9.
Often, it is of interest to characterize the variation of F(θ , φ )
in the form of two-dimensional plots in specific planes in the
spherical coordinate system. The two planes most commonly
specified for this purpose are the elevation and azimuth planes.
The elevation plane, also called the θ plane, is a plane
corresponding to a constant value of φ . For example, φ = 0
defines the x–z plane and φ = 90◦ defines the y–z plane, both
of which are elevation planes (Fig. 9-8). A plot of F(θ , φ )
versus θ in either of these planes constitutes a two-dimensional
pattern in the elevation plane. This is not to imply, however,
that the elevation-plane pattern is necessarily the same in all
elevation planes.
The azimuth plane, also called the φ plane, is specified by
θ = 90◦ and corresponds to the x–y plane. The elevation and
azimuth planes are often called the two principal planes of the
spherical coordinate system.
Some antennas exhibit highly directive patterns with narrow
beams, in which case it is often convenient to plot the antenna
pattern on a decibel scale by expressing F in decibels:
−5
th
ni
Ze
9-2.1 Antenna Pattern
Normalized radiation intensity (dB)
Prad = R2
Z 2π Z π
A
Figure 9-9
eφ
ngl
a
uth
zim
−1
92 −2
es)
gre
e
d
(
Three-dimensional pattern of a narrow-beam
antenna.
Another format commonly used for inspecting the pattern
of a narrow-beam antenna is the rectangular display shown in
Fig. 9-10(b), which permits the pattern to be easily expanded
by changing the scale of the horizontal axis. These plots
represent the variation in only one plane in the observation
sphere: the φ = 0 plane. Unless the pattern is symmetrical in φ ,
additional patterns are required to define the overall variation
of F(θ , φ ) with θ and φ .
Strictly speaking, the polar angle θ is always positive,
being defined over the range from 0◦ (z direction) to 180◦
(−z direction), yet the θ axis in Fig. 9-10(b) is shown to have
both positive and negative values. This is not a contradiction,
but rather a different form of plotting antenna patterns. The
right-hand half of the plot represents the variation of F (dB)
with θ because θ is increased in a clockwise direction in the
x–z plane [see inset in Fig. 9-10(b)], corresponding to φ = 0.
The left-hand half of the plot represents the variation of F (dB)
with θ as θ is increased in a counterclockwise direction at
φ = 180◦. Thus, a negative θ value simply denotes that the
direction (θ , φ ) is in the left-hand half of the x–z plane.
The pattern shown in Fig. 9-10(a) indicates that the antenna
is fairly directive, since most of the energy is radiated through
9-2 ANTENNA RADIATION CHARACTERISTICS
−10
30
Zenith angle θ (degrees)
−20
40
−30
50
60
10
20
0
30
First side lobe
40
Minor
lobes
50
60
−40
70
70
80
80
90
90
100
100
110
110
Back
lobes
120
−3
−5
Main lobe
Normalized radiation intensity F(θ), dB
0
Normalized radiation intensity, dB
10
20
393
β1/2
θ = 0°
z
θ
−10
φ = 180°
x
θ = 90°
φ = 0°
θ = 180°
−15
−20
−25
βnull
−30
120
130
130
140 150 160 170 180 170 160 150 140
(a) Polar diagram
−35
−50 −40 −30 −20 −10 θ1 0 θ2 10 20 30
Zenith angle θ (degrees)
40
50
(b) Rectangular plot
Figure 9-10 Representative plots of the normalized radiation pattern of a microwave antenna in (a) polar form and (b) rectangular form.
a narrow sector called the main lobe. In addition to the
main lobe, the pattern exhibits several side lobes and back
lobes as well. For most applications, these extra lobes are
considered undesirable because they represent wasted energy
for transmitting antennas and potential interference directions
for receiving antennas.
1
F(θ, φ)
1
F = 1 within
the cone
Ωp
9-2.2 Beam Dimensions
For an antenna with a single main lobe, the pattern solid angle
Ωp describes the equivalent width of the main lobe of the
antenna pattern (Fig. 9-11). It is defined as the integral of the
normalized radiation intensity F(θ , φ ) over a sphere:
Ωp =
ZZ
4π
F(θ , φ ) dΩ
(sr).
(9.21)
(a) Actual pattern
(b) Equivalent solid angle
Figure 9-11 The pattern solid angle Ωp defines an equivalent
cone over which all of the radiation of the actual antenna is
concentrated with uniform intensity equal to the maximum of
the actual pattern.
394
CHAPTER 9 RADIATION AND ANTENNAS
◮ For an isotropic antenna with F(θ , φ ) = 1 in all directions, Ωp = 4π (sr). ◭
z
The pattern solid angle characterizes the directional properties of the three-dimensional radiation pattern. To characterize
the width of the main lobe in a given plane, the term used
is beamwidth. The half-power beamwidth, or simply the
beamwidth β , is defined as the angular width of the main lobe
between the two angles at which the magnitude of F(θ , φ ) is
equal to half of its peak value (or −3 dB on a decibel scale).
For example, for the pattern displayed in Fig. 9-10(b), β is
given by
(9.22)
β = θ2 − θ1 ,
where θ1 and θ2 are the half-power angles at which
F(θ , 0) = 0.5 (with θ2 denoting the larger value and θ1
denoting the smaller one, as shown in the figure). If the pattern
is symmetrical and the peak value of F(θ , φ ) is at θ = 0,
then β = 2θ2 . For the short-dipole pattern shown earlier in
Fig. 9-7(a), F(θ ) is maximum at θ = 90◦ , θ2 is at 135◦, and
θ1 is at 45◦ . Hence, β = 135◦ − 45◦ = 90◦ . The beamwidth β
is also known as the 3 dB beamwidth. In addition to the halfpower beamwidth, other beam dimensions may be of interest
for certain applications, such as the null beamwidth βnull ,
which is the angular width between the first nulls on the two
sides of the peak (Fig. 9-10(b)).
9-2.3 Antenna Directivity
The directivity D of an antenna is defined as the ratio of
its maximum normalized radiation intensity, Fmax (which by
definition is equal to 1), to the average value of F(θ , φ ) over
all directions (4π space):
D=
1
Fmax
4π
ZZ
=
.
=
1
Fav
Ωp
F(θ , φ ) dΩ
4 π 4π
(9.23)
(dimensionless)
Here Ωp is the pattern solid angle defined by Eq. (9.21).
Thus, the narrower Ωp of an antenna pattern is, the greater is
the directivity. For an isotropic antenna, Ωp = 4π ; hence, its
directivity Diso = 1.
By using Eq. (9.20) in Eq. (9.23), D can be expressed as
4π R2 Smax Smax
=
,
D=
Prad
Sav
0 dB
βxz
βyz
y
x
Figure 9-12 The solid angle of a unidirectional radiation
pattern is approximately equal to the product of the half-power
beamwidths in the two principal planes; that is, Ωp ≈ βxz βyz .
the antenna, Prad , divided by the surface area of a sphere of
radius R.
◮ Since Sav = Siso , where Siso is the power density radiated
by an isotropic antenna, D represents the ratio of the
maximum power density radiated by the antenna to the
power density radiated by an isotropic antenna, where
both are measured at the same range R and excited by the
same amount of input power. ◭
Usually, D is expressed in decibels:∗ D (dB) = 10 log D.
For an antenna with a single main lobe pointing in the z
direction shown in Fig. 9-12, Ωp may be approximated as the
product of the half-power beamwidths βxz and βyz (in radians):
Ωp ≈ βxz βyz ,
and
D=
(9.24)
where Sav = Prad /(4π R2) is the average value of the radiated
power density and is equal to the total power radiated by
(9.25)
4π
4π
.
≈
Ωp
βxz βyz
(single main lobe)
(9.26)
∗ A note of caution: Even though we often express certain dimensionless
quantities in decibels, we should always convert their decibel values to natural
values before using them in the relations given in this chapter.
9-2 ANTENNA RADIATION CHARACTERISTICS
Although approximate, this relation provides a useful method
for estimating the antenna directivity from measurements of
the beamwidths in the two orthogonal planes whose intersection is the axis of the main lobe.
Example 9-1:
Antenna Radiation Properties
Determine (a) the direction of maximum radiation, (b) pattern
solid angle, (c) directivity, and (d) half-power beamwidth
in the y–z plane for an antenna that radiates only into the
upper hemisphere with normalized radiation intensity given by
F(θ , φ ) = cos2 θ .
Solution: The statement that the antenna radiates through
only the upper hemisphere is equivalent to

 cos2 θ for 0 ≤ θ ≤ π /2
and 0 ≤ φ ≤ 2π ,
F(θ , φ ) = F(θ ) =
 0
elsewhere.
(a) The function F(θ ) = cos2 θ is independent of φ and is
maximum when θ = 0◦ . A polar plot of F(θ ) is shown in
Fig. 9-13.
1
0.5
Figure 9-13 Polar plot of F(θ ) = cos2 θ .
(b) From Eq. (9.21), the pattern solid angle Ωp is given by
ZZ
Z 2π Z π /2
2
Ωp =
F(θ , φ ) dΩ =
cos θ sin θ d θ d φ
θ =0
=
=
Z 2π φ =0
Z 2π
1
0
cos3 θ
−
3
which gives the half-power angles θ1 = −45◦ and θ2 = 45◦ .
Hence,
β = θ2 − θ1 = 90◦ .
Example 9-2: Directivity of a Hertzian Dipole
Calculate the directivity of a Hertzian dipole.
Solution: Application of Eq. (9.23) with F(θ ) = sin2 θ [from
Eq. (9.15)] gives
4π
2π
F(θ , φ ) sin θ d θ d φ
Z π
φ =0 θ =0
4π
=
sin3 θ d θ d φ
4π
= 1.5
8π /3
9-2.4 Antenna Gain
x
φ =0
F(θ ) = cos2 θ = 0.5,
or, equivalently, 1.76 dB.
y
4π
(d) The half-power beamwidth β is obtained by setting
F(θ ) = 0.5. That is,
=Z
45°
90°
which corresponds to D (dB) = 10 log6 = 7.78 dB.
4π
F(θ) = cos2 θ
0.5
(c) Application of Eq. (9.23) gives
4π
3
= 4π
D=
= 6,
Ωp
2π
D = ZZ
z
–45°
395
π /2
2π
dφ =
3
3
dφ
0
(sr).
Of the total power Pt (transmitter power) supplied to the
antenna, a part, Prad , is radiated out into space, and the
remainder, Ploss , is dissipated as heat in the antenna structure.
The radiation efficiency ξ is defined as the ratio of Prad to Pt :
ξ=
Prad
.
Pt
(dimensionless)
(9.27)
The gain of an antenna is defined as
G=
4π R2Smax
,
Pt
(9.28)
which is similar in form to the expression given by Eq. (9.24)
for the directivity D except that it is referenced to the input
396
CHAPTER 9 RADIATION AND ANTENNAS
power supplied to the antenna, Pt , rather than to the radiated
power Prad . In view of Eq. (9.27),
G = ξ D.
(dimensionless)
(9.29)
◮ The gain accounts for ohmic losses in the antenna
material, whereas the directivity does not. For a lossless
antenna, ξ = 1, and G = D. ◭
9-2.5 Radiation Resistance
To a transmission line connected between a generator supplying power Pt on one end and an antenna on the other end, the
antenna is merely a load with input impedance Zin . If the line
is lossless and properly matched to the antenna, all of Pt is
transferred to the antenna. In general, Zin consists of a resistive
component Rin and a reactive component Xin :
Zin = Rin + jXin .
(9.30)
The resistive component is defined as equivalent to a resistor Rin that would consume an average power Pt when the
amplitude of the ac current flowing through it is I0 ,
Pt =
1 2
2 I0 Rin .
(9.31)
Since Pt = Prad + Ploss , it follows that Rin can be defined as the
sum of a radiation resistance Rrad and a loss resistance Rloss ,
Rin = Rrad + Rloss ,
Example 9-3:
Radiation Resistance and
Efficiency of a Hertzian Dipole
A 4-cm long center-fed dipole is used as an antenna at
75 MHz. The antenna wire is made of copper and has a radius
a = 0.4 mm. From Eqs. (7.92a) and (7.94), the loss resistance
of a circular wire of length l is given by
r
π f µc
l
,
(9.35)
Rloss =
2π a
σc
where µc and σc are the magnetic permeability and conductivity of the wire, respectively. Calculate the radiation resistance
and the radiation efficiency of the dipole antenna.
Solution: At 75 MHz,
λ=
3 × 108
c
= 4 m.
=
f
7.5 × 107
The length to wavelength ratio is l/λ = 4 cm/4 m = 10−2.
Hence, this is a short dipole. From Eq. (9.24),
4 π R2
Smax .
(9.36)
D
For the Hertzian dipole, Smax is given by Eq. (9.14), and from
Example 9-2, we established that D = 1.5. Hence,
2
l
4π R2 15π I02 l 2
2 2
Prad =
×
= 40π I0
.
(9.37)
2
1.5
R
λ
λ
Prad =
Equating this result to Eq. (9.33a) and then solving for the
radiation resistance Rrad leads to
(9.32)
Rrad = 80π 2(l/λ )2
with
Prad =
Ploss =
1 2
2 I0 Rrad ,
1 2
2 I0 Rloss ,
(9.33a)
(9.33b)
where I0 is the amplitude of the sinusoidal current exciting the
antenna. As defined earlier, the radiation efficiency is the ratio
of Prad to Pt , or
ξ=
Prad
Rrad
Prad
=
=
.
Pt
Prad + Ploss Rrad + Rloss
(9.34)
The radiation resistance Rrad can be calculated by integrating
the far-field power density over a sphere to obtain Prad and then
equating the result to Eq. (9.33a).
(Ω). (short dipole)
(9.38)
For l/λ = 10−2 , Rrad = 0.08 Ω.
Next, we determine the loss resistance Rloss . For copper, Appendix B gives µc ≈ µ0 = 4π × 10−7 H/m and
σc = 5.8 × 107 S/m. Hence,
r
π f µc
l
Rloss =
2π a
σc
1/2
π × 75 × 106 × 4π × 10−7
4 × 10−2
=
2π × 4 × 10−4
5.8 × 107
= 0.036 Ω.
Therefore, the radiation efficiency is
ξ=
0.08
Rrad
=
= 0.69.
Rrad + Rloss 0.08 + 0.036
9-3 HALF-WAVE DIPOLE ANTENNA
397
Thus, the dipole is 69% efficient.
Concept Question 9-5:
Current distrubution
I(z) = I0 cos kz
What does the pattern solid
angle represent?
What is the magnitude of the
directivity of an isotropic antenna?
Concept Question 9-6:
l = λ/2
What physical and material
properties affect the radiation efficiency of a fixed-length
Hertzian dipole antenna?
Concept Question 9-7:
Transmission
line
i(t)
Dipole
antenna
i(t)
(a)
Exercise 9-2: An antenna has a conical radiation pattern
z
with a normalized radiation intensity F(θ ) = 1 for θ
between 0◦ and 45◦ and zero for θ between 45◦ and 180◦ .
The pattern is independent of the azimuth angle φ . Find
(a) the pattern solid angle and (b) the directivity.
Answer: (a) Ωp = 1.84 sr, (b) D = 6.83 or, equivalently,
8.3 dB. (See
EM
.)
z = l/2
dz
z
Q = (R, θ, φ)
s
θs
R
θ
l = λ/2
z cos θ
Exercise 9-3: The maximum power density radiated by
a short dipole at a distance of 1 km is 60 (nW/m2 ). If
I0 = 10 A, find the radiation resistance.
Answer: Rrad = 10 mΩ. (See
EM
z = –l/2
(b)
.)
Figure 9-14 Center-fed half-wave dipole.
9-3 Half-Wave Dipole Antenna
In Section 9-1, we developed expressions for the electric
and magnetic fields radiated by a Hertzian dipole of length
l ≪ λ . We now use these expressions as building blocks to
obtain expressions for the fields radiated by a half-wave dipole
antenna, so named because its length l = λ /2. As shown in
Fig. 9-14, the half-wave dipole consists of a thin wire fed at
its center by a generator connected to the antenna terminals
via a transmission line. The current flowing through the wire
has a symmetrical distribution with respect to the center of the
dipole, and the current is zero at its ends. Mathematically, i(t)
is given by
i(t) = I0 cos ω t cos kz = Re I0 coskz e jω t ,
(9.39a)
whose phasor is
e = I0 coskz,
I(z)
−λ /4 ≤ z ≤ λ /4 ,
(9.39b)
and k = 2π /λ . Equation (9.9a) gives an expression for Eeθ ,
which is the far field radiated by a Hertzian dipole of length l
when excited by a current I0 . Let us adapt that expression to an
infinitesimal dipole segment of length dz, which is excited by
e and located at a distance s from the observation
a current I(z)
point Q (Fig. 9-14(b)). Thus,
− jks jkη0 e
e
e
sin θs ,
I(z) dz
(9.40a)
d Eθ (z) =
4π
s
and the associated magnetic field is
eφ (z) =
dH
d Eeθ (z)
.
η0
(9.40b)
The far field due to radiation by the entire antenna is obtained
by integrating the fields from all of the Hertzian dipoles
making up the antenna:
Eeθ =
Z λ /4
z=−λ /4
d Eeθ .
(9.41)
Before we calculate this integral, we make the following
two approximations. The first relates to the magnitude part
398
CHAPTER 9 RADIATION AND ANTENNAS
of the spherical propagation factor, 1/s. In Fig. 9-14(b), the
distance s between the current element and the observation
point Q is considered so large in comparison with the length of
the dipole that the difference between s and R may be neglected
in terms of its effect on 1/s. Hence, we may set 1/s ≈ 1/R,
and by the same argument, we set θs ≈ θ . The error ∆ between
s and R is a maximum when the observation point is along
the z axis and it is equal to λ /4 (corresponding to half of the
antenna length). If R ≫ λ , this error will have an insignificant
effect on 1/s. The second approximation is associated with the
phase factor e− jks . An error in distance ∆ corresponds to an
error in phase k∆ = (2π /λ )(λ /4) = π /2. As a rule of thumb,
a phase error greater than π /8 is considered unacceptable
because it may lead to a significant error in the computed value
of the field Eeθ . Hence, the approximation s ≈ R is too crude for
the phase factor and cannot be used. A more tolerable option
is to use the parallel-ray approximation given by
s ≈ R − z cos θ ,
(9.42)
as illustrated in Fig. 9-14(b).
Substituting Eq. (9.42) for s in the phase factor of
Eq. (9.40a) and replacing s with R and θs with θ elsewhere
in the expression, we obtain
− jkR jkη0 e
e
sin θ e jkz cos θ .
I(z) dz
d Eeθ =
(9.43)
4π
R
After (1) inserting Eq. (9.43) into Eq. (9.41), (2) using the
e given by Eq. (9.39b), and (3) carrying out
expression for I(z)
the integration, the following expressions are obtained:
Eeθ = j 60I0
cos[(π /2) cos θ ]
sin θ
e
eφ = Eθ .
H
η0
e− jkR
R
,
(9.44a)
(9.44b)
Hence, the normalized radiation intensity is
S(R, θ )
=
S0
F(θ ) =
cos[(π /2) cos θ ]
sin θ
2
.
(9.46)
The radiation pattern of the half-wave dipole exhibits roughly
the same doughnut-like shape shown earlier in Fig. 9-7 for the
short dipole. Its directivity is slightly larger (1.64 compared
with 1.5 for the short dipole), but its radiation resistance is
73 Ω (as shown later in Section 9-3.2), which is orders of
magnitude larger than that of a short dipole.
9-3.1
Directivity of λ /2 Dipole
To evaluate both the directivity D and the radiation resistance
Rrad of the half-wave dipole, we first need to calculate the total
radiated power Prad by applying Eq. (9.20):
Prad = R2
ZZ
4π
15I02
=
π
S(R, θ ) dΩ
Z 2πZ π cos[(π /2) cos θ ] 2
0
0
sin θ
sin θ d θ d φ .
(9.47)
The integration over φ is equal to 2π , and numerical evaluation
of the integration over θ gives the value 1.22. Consequently,
Prad = 36.6 I02
(W).
(9.48)
From Eq. (9.45), we found that Smax = 15I02 /(π R2). Using
this in Eq. (9.24) gives the following result for the directivity D
of the half-wave dipole:
4π R2 15I02
4π R2Smax
= 1.64
(9.49)
=
D=
Prad
36.6I02 π R2
or, equivalently, 2.15 dB.
The corresponding time-average power density is
15I02 cos2 [(π /2) cos θ ]
|Eeθ |2
S(R, θ ) =
=
2η0
π R2
sin2 θ
2
cos [(π /2) cos θ ]
= S0
(W/m2 ).
sin2 θ
(9.45)
Examination of Eq. (9.45) reveals that S(R, θ ) is maximum at
θ = π /2, and its value is
Smax = S0 =
15I02
.
π R2
9-3.2
Radiation Resistance of λ /2 Dipole
From Eq. (9.33a),
Rrad =
2Prad 2 × 36.6I02
=
≈ 73 Ω.
I02
I02
(9.50)
As was noted earlier in Example 9-3, because the radiation
resistance of a Hertzian dipole is comparable in magnitude
to that of its loss resistance Rloss , its radiation efficiency ξ
is rather small. For the 4-cm long dipole of Example 9-3,
Rrad = 0.08 Ω (at 75 MHz) and Rloss = 0.036 Ω. If we keep
9-3 HALF-WAVE DIPOLE ANTENNA
399
the frequency the same and increase the length of the dipole to
2 m (λ = 4 m at f = 75 MHz), Rrad becomes 73 Ω and Rloss
increases to 1.8 Ω. The radiation efficiency increases from
69% for the short dipole to 98% for the half-wave dipole. More
significant is the fact that it is practically impossible to match
a transmission line to an antenna with a resistance on the order
of 0.1 Ω, while it is quite easy to do so when Rrad = 73 Ω.
Moreover, since Rloss ≪ Rrad for the half-wave dipole,
Rin ≈ Rrad and Eq. (9.30) becomes
Zin ≈ Rrad + jXin .
λ/4
I
(9.51)
Deriving an expression for Xin for the half-wave dipole is fairly
complicated and beyond the scope of this book. However,
it is significant to note that Xin is a strong function of l/λ
and that it decreases from 42 Ω at l/λ = 0.5 to zero at
l/λ = 0.48, whereas Rrad remains approximately unchanged.
Hence, by reducing the length of the half-wave dipole by 4%,
Zin becomes purely real and equal to 73 Ω, thereby making
it possible to match the dipole to a 75 Ω transmission line
without resorting to the use of a matching network.
Conducting
plane
(a)
I
I
Image
9-3.3 Quarter-Wave Monopole Antenna
(b)
◮ When placed over a conducting ground plane, a quarterwave monopole antenna excited by a source at its base
(Fig. 9-15(a)) exhibits the same radiation pattern in the
region above the ground plane as a half-wave dipole in
free space. ◭
Figure 9-15 A quarter-wave monopole above a conducting
plane is equivalent to a full half-wave dipole in free space.
How does the radiation pattern
of a half-wave dipole compare with that of a Hertzian
dipole? How do their directivities, radiation resistances,
and radiation efficiencies compare?
Concept Question 9-9:
This is because, from image theory (Section 4-11), the conducting plane can be replaced with the image of the λ /4
monopole, as illustrated in Fig. 9-15(b). Thus, the λ /4
monopole radiates an electric field identical to that given by
Eq. (9.44a), and its normalized radiation intensity is given
by Eq. (9.46); but the radiation is limited to the upper halfspace defined by 0 ≤ θ ≤ π /2. Hence, a monopole radiates
only half as much power as the dipole. Consequently, for a
λ /4 monopole, Prad = 18.3I02, and its radiation resistance is
Rrad = 36.5 Ω.
The approach used with the quarter-wave monopole is also
valid for any vertical wire antenna placed above a conducting
plane, including a Hertzian monopole.
What is the physical length of
a half-wave dipole operating at (a) 1 MHz (in the AM
broadcast band), (b) 100 MHz (FM broadcast band), and
(c) 10 GHz (microwave band)?
Concept Question 9-8:
How does the radiation efficiency of a quarter-wave monopole compare with that of
a half-wave dipole, assuming that both are made of the
same material and have the same cross section?
Concept Question 9-10:
Exercise 9-4: For the half-wave dipole antenna, evaluate
F(θ ) versus θ to determine the half-power beamwidth in
the elevation plane (the plane containing the dipole axis).
Answer: β = 78◦ . (See
EM
.)
400
CHAPTER 9 RADIATION AND ANTENNAS
Exercise 9-5: If the maximum power density radiated by
µ W/m2
a half-wave dipole is 50
is the current amplitude I0 ?
Answer: I0 = 3.24 A. (See
EM
at a range of 1 km, what
.)
9-4 Dipole of Arbitrary Length
So far, we examined the radiation properties of the Hertzian
and half-wave dipoles. We now consider the more general case
of a linear dipole of arbitrary length l relative to λ . For a
center-fed dipole, as depicted in Fig. 9-16, the currents flowing
through its two halves are symmetrical and must go to zero at
e can be expressed as a
its ends. Hence, the current phasor I(z)
sine function with an argument that goes to zero at z = ±l/2:
(
e = I0 sin [k (l/2 − z)] , for 0 ≤ z ≤ l/2,
I(z)
(9.52)
I0 sin [k (l/2 + z)] , for − l/2 ≤ z < 0,
where I0 is the current amplitude. The procedure for calculating the electric and magnetic fields and the associated
power density of the wave radiated by such an antenna is
basically the same as that used previously in connection with
the half-wave dipole antenna. The only difference is the current
e
e given by
distribution I(z).
If we insert the expression for I(z)
~
I(z)
~
I(z)
The total field radiated by the dipole is
Eeθ =
Z l/2
−l/2
d Eeθ =
Z l/2
0
d Eeθ +
Z 0
−l/2
d Eeθ
jkη0 I0 e− jkR
=
sin θ
4π
R
Z l/2
×
e jkz cos θ sin[k(l/2 − z)] dz
0
+
Z 0
−l/2
e
jkz cos θ
sin[k(l/2 + z)] dz .
(9.54)
If we apply Euler’s identity to express e jkz cos θ as
e jkz cos θ = cos(kz cos θ ) + j sin(kz cos θ )],
we can integrate the two integrals and obtain the result
#
− jkR "
cos kl2 cos θ − cos kl2
e
Eeθ = j60I0
. (9.55)
R
sin θ
The corresponding time-average power density radiated by the
dipole antenna is given by
"
#2
15I02 cos πλl cos θ − cos πλl
|Eeθ |2
S(θ ) =
=
, (9.56)
2η0
π R2
sin θ
~
I(z)
(a) l = λ/2
Eq. (9.52) into Eq. (9.43), we obtain the following expression
for the differential electric field d Eeθ of the wave radiated by
an elemental length dz at location z along the dipole:
jkη0 I0 e− jkR
e
sin θ e jkz cos θ dz
d Eθ =
4π
R
(
sin [k (l/2 − z)]
for 0 ≤ z ≤ l/2,
(9.53)
×
sin [k (l/2 + z)]
for − l/2 ≤ z < 0.
(b) l = λ
(c) l = 3λ/2
Figure 9-16 Current distribution for three center-fed dipoles.
where we have used the relations η0 ≈ 120π (Ω) and
k = 2π /λ . For l = λ /2, Eq. (9.56) reduces to the expression given by Eq. (9.45) for the half-wave dipole. Plots of
the normalized radiation intensity, F(θ ) = S(R, θ )/Smax, are
shown in Fig. 9-17 for dipoles of lengths λ /2, λ , and 3λ /2.
The dipoles with l = λ /2 and l = λ have similar radiation
patterns with both maxima along θ = 90◦ , but the half-power
beamwidth of the wavelength-long dipole is narrower than
that of the half-wave dipole, and Smax = 60I02/(π R2 ) for the
wavelength-long dipole, which is four times that for the halfwave dipole. The pattern of the dipole with length l = 3λ /2
exhibits a structure with multiple lobes, and its direction of
maximum radiation is not along θ = 90◦ .
9-5 EFFECTIVE AREA OF A RECEIVING ANTENNA
401
9-5 Effective Area of a Receiving
Antenna
z
0.5
β = 78°
1
x–y plane
So far, antennas have been treated as directional radiators of
energy. Now, we examine the reverse process, namely how a
receiving antenna extracts energy from an incident wave and
delivers it to a load. The ability of an antenna to capture energy
from an incident wave of power density Si (W/m2 ) and convert
it into an intercepted power Pint (W) for delivery to a matched
load is characterized by the effective area Ae :
(a) l = λ/2
Ae =
z
0.5
β = 47°
1
x–y plane
(b) l = λ
z
Pint
Si
(m2 ).
(9.57)
Other commonly used names for Ae include effective aperture
and receiving cross section. The antenna receiving process
may be modeled in terms of a Thévenin equivalent circuit
eoc in series with the
(Fig. 9-18) consisting of a voltage V
antenna input impedance Zin . Here, Veoc is the open-circuit
voltage induced by the incident wave at the antenna terminals,
and ZL is the impedance of the load connected to the antenna
(representing a receiver or some other circuit). In general, both
Zin and ZL are complex:
Zin = Rrad + jXin ,
(9.58a)
x–y plane
Incident
wave
ZL
Load
(c) l = 3λ/2
Antenna
(a) Receiving antenna
Figure 9-17 Radiation patterns of dipoles with lengths of λ /2,
λ , and 3λ /2.
Zin = Rrad + jXin
Module 9.2 Linear Dipole Antenna For a linear
antenna of any specified length (in units of λ ), this
module displays the current distribution along the antenna
and the far-field radiation patterns in the horizontal and
elevation planes. It also calculates the total power radiated
by the antenna, the radiation resistance, and the antenna
directivity.
~
Voc
ZL = RL + jXL
Antenna equivalent circuit
Load
(b) Equivalent circuit
Figure 9-18 Receiving antenna represented by an equivalent
circuit.
402
CHAPTER 9 RADIATION AND ANTENNAS
Module 9.3 Detailed Analysis of Linear Antenna This module complements Module 9.2 by offering extensive
information about the specified linear antenna, including its directivity and plots of its current and field distributions.
ZL = RL + jXL ,
(9.58b)
where Rrad denotes the radiation resistance of the antenna
(assuming Rloss ≪ Rrad ). To maximize power transfer to
the load, the load impedance must be chosen so that either
ZL = Zin∗ , or equivalently, RL = Rrad and XL = −Xin . In that
eoc that is connected across
case, the circuit reduces to a source V
a resistance equal to 2Rrad . Since Veoc is a sinusoidal voltage
phasor, the time-average power delivered to the load is
"
#2
|Veoc |2
1 |Veoc |
1 e 2
Rrad =
,
PL = |IL | Rrad =
2
2 2Rrad
8Rrad
(9.59)
eoc /(2Rrad ) is the phasor current flowing through
where IeL = V
the circuit. Since the antenna is lossless, all of the intercepted
9-6 FRIIS TRANSMISSION FORMULA
403
power Pint ends up in the load resistance RL . Hence,
Pint = PL =
|Veoc |2
.
8Rrad
(9.60)
Answer: (a) D = 9.67, (b) Ae = 6.92 m2 . (See
For an incident wave with electric field Eei parallel to the
antenna polarization direction, the power density carried by
the wave is
|Eei |2
|Eei |2
=
.
(9.61)
Si =
2η0
240π
The ratio of the results provided by Eqs. (9.60) and (9.61) gives
Ae =
Pint
30π |Veoc|2
.
=
Si
Rrad |Eei |2
(9.62)
eoc induced in the receiving antenna
The open-circuit voltage V
is due to the incident field Eei , but the relation between them
depends on the specific antenna under consideration. By way
of illustration, let us consider the case of the short-dipole
antenna of Section 9-1. Because the length l of the short dipole
is small compared with λ , the current induced by the incident
field is uniform across its length, and the open-circuit voltage
is simply Veoc = Eei l. Noting that Rrad = 80π 2(l/λ )2 for the
short dipole (see Eq. (9.38)) and using Veoc = Eei l, Eq. (9.62)
simplifies to
Ae =
3λ 2
8π
(m2 ).
(short dipole)
λ 2D
4π
(m2 ). (any antenna)
(9.64)
◮ Despite the fact that the relation between Ae and D
given by Eq. (9.64) was derived for a Hertzian dipole, it
can be shown that it is also valid for any antenna under
matched-impedance conditions. ◭
Exercise 9-6: The effective area of an antenna is 9 m2 .
EM
.)
9-6 Friis Transmission Formula
The two antennas shown in Fig. 9-19 are part of a freespace communication link with the separation between the
antennas, R, being large enough for each to be in the far-field
region of the other. The transmitting and receiving antennas
have effective areas At and Ar and radiation efficiencies ξt
and ξr , respectively. Our objective is to find a relationship
between Pt , which is the power supplied to the transmitting
antenna, and Prec , which is the power delivered to the receiver.
As always, we assume that both antennas are impedancematched to their respective transmission lines. Initially, we
consider the case where the two antennas are oriented so that
the peak of the radiation pattern of each antenna points in the
direction of the other.
We start by treating the transmitting antenna as a lossless
isotropic radiator. The power density incident upon the receiving antenna at a distance R from an isotropic transmitting
antenna is simply equal to the transmitter power Pt divided by
the surface area of a sphere of radius R:
Siso =
(9.63)
In Example 9-2 it was shown that for the Hertzian dipole the
directivity D = 1.5. In terms of D, Eq. (9.63) can be rewritten
in the form
Ae =
Exercise 9-7: At 100 MHz, the pattern solid angle of an
antenna is 1.3 sr. Find (a) the antenna directivity D and (b)
its effective area Ae .
Pt
.
4 π R2
(9.65)
The real transmitting antenna is neither lossless nor isotropic.
Hence, the power density Sr due to the real antenna is
Sr = Gt Siso = ξt Dt Siso =
ξt Dt Pt
.
4 π R2
Through the gain Gt = ξt Dt , ξt accounts for the fact that only
part of the power Pt supplied to the antenna is radiated out into
space, and Dt accounts for the directivity of the transmitting
Ar
At
R
Pt
Prad
Transmitting
antenna
Prec
Pint
Receiving
antenna
What is its directivity in decibels at 3 GHz?
Answer: D = 40.53 dB. (See
EM
.)
(9.66)
Figure 9-19 Transmitter–receiver configuration.
404
CHAPTER 9 RADIATION AND ANTENNAS
antenna (in the direction of the receiving antenna). Moreover,
by Eq. (9.64), Dt is related to At by Dt = 4π At /λ 2 . Hence,
Eq. (9.66) becomes
ξt At Pt
Sr = 2 2 .
(9.67)
λ R
On the receiving-antenna side, the power intercepted by the
receiving antenna is equal to the product of the incident power
density Sr and the effective area Ar :
Pint = Sr Ar =
ξt At Ar Pt
.
λ 2 R2
(9.68)
The power delivered to the receiver, Prec , is equal to the
intercepted power Pint multiplied by the radiation efficiency of
the receiving antenna, ξ r . Hence, Prec = ξr Pint , which leads to
the result
Prec
ξt ξr A t A r
=
= Gt Gr
Pt
λ 2 R2
λ
4π R
2
.
performance of the receiver–antenna combination, K is Boltzmann’s constant [1.38 × 10−23 (J/K)], and B is the receiver
bandwidth in Hz.
The signal-to-noise ratio Sn (which should not be confused
with the power density S) is defined as the ratio of Prec to Pn :
Sn = Prec /Pn
(9.72)
For a receiver with Tsys = 580 K, what minimum diameter of a
parabolic dish receiving antenna is required for high-quality
TV reception with Sn = 40 dB? The satellite and ground
receiving antennas may be assumed lossless, and their effective
areas may be assumed equal to their physical apertures.
Solution: The following quantities are given:
Pt = 100 W,
f = 6 GHz = 6 × 109 Hz,
Sn = 104 ,
transmit antenna diameter dt = 2 m,
Tsys = 580 K,
(9.69)
◮ This relation is known as the Friis transmission formula, and Prec /Pt is called the power transfer ratio. ◭
(dimensionless).
R = 40, 000 km = 4 × 107 m,
B = 5 MHz = 5 × 106 Hz.
The wavelength λ = c/ f = 5 × 10−2 m, and the area of the
transmitting satellite antenna is At = (π dt2 /4) = π (m2 ). From
Eq. (9.71), the receiver noise power is
Pn = KTsys B = 1.38 × 10−23 × 580 × 5 × 106 = 4 × 10−14 W.
If the two antennas are not oriented in the direction of
maximum power transfer, Eq. (9.69) assumes the general form
Prec
= Gt Gr
Pt
λ
4π R
2
Prec =
Ft (θt , φt ) Fr (θr , φr ),
(9.70)
where Ft (θt , φt ) is the normalized radiation intensity of the
transmitting antenna at angles (θ t , φt ) corresponding to the
direction of the receiving antenna (as seen by the antenna
pattern of the transmitting antenna), and a similar definition
applies to Fr (θr , φr ) for the receiving antenna.
Example 9-4:
Using Eq. (9.69) with ξt = ξr = 1,
Satellite Communication System
A 6 GHz direct-broadcast TV satellite system transmits 100 W
through a 2 m diameter parabolic dish antenna from a distance
of approximately 40,000 km above Earth’s surface. Each TV
channel occupies a bandwidth of 5 MHz. Due to electromagnetic noise picked up by the antenna as well as noise generated
by the receiver electronics, a home TV receiver has a noise
level given by
Pn = KTsys B
(W),
(9.71)
where Tsys [measured in kelvins (K)] is a figure of merit called
the system noise temperature that characterizes the noise
Pt At Ar
100π Ar
= 7.85 × 10−11Ar .
=
λ 2 R2
(5 × 10−2)2 (4 × 107)2
The area of the receiving antenna, Ar , can now be determined
by equating the ratio Prec /Pn to Sn = 104 :
104 =
7.85 × 10−11Ar
,
4 × 10−14
which yields thep
value Ar = 5.1 m2 . The required minimum
diameter is dr = 4Ar /π = 2.55 m.
Exercise 9-8: If the operating frequency of the com-
munication system described in Example 9-4 is doubled
to 12 GHz, what would then be the minimum required
diameter of a home receiving TV antenna?
Answer: dr = 1.27 m. (See
EM
.)
Exercise 9-9: A 3 GHz microwave link consists of two
identical antennas each with a gain of 30 dB. Determine
the received power given that the transmitter output power
is 1 kW and the two antennas are 10 km apart.
Answer: Prec = 6.33 × 10−4 W. (See
EM
.)
9-6 FRIIS TRANSMISSION FORMULA
Technology Brief 17:
Health Risks of EM Fields
Can the use of cell phones cause cancer? Does exposure to the electromagnetic fields (EMFs) associated
with power lines pose health risks to humans? Are we
endangered by EMFs generated by home appliances,
telephones, electrical wiring, and the myriad of electronic gadgets we use every day (Fig. TF17-1)? Despite
reports in some of the popular media alleging a causative
relationship between low-level EMFs and many diseases, according to reports issude by governmental and
professional boards in the U.S. and Europe, the answer
is given here.
◮ NO, we are not at risk, so long as manufacturers
adhere to the approved governmental standards
for maximum permissible exposure (MPE) levels.
With regard to cell phones, the official reports caution that their conclusions are limited to phone use
of less than 15 years, since data for longer use is not
yet available. ◭
Figure TF17-1 Electromagnetic fields are emitted by power lines,
cell phones, TV towers, and many other electronic circuits and
devices.
405
Physiological Effects of EMFs
The energy carried by a photon with an EM frequency f
is given by E = h f , where h is Planck’s constant. The
mode of interaction between a photon passing through
a material and the material’s atoms or molecules is
very much dependent on f . If f is greater than about
1015 Hz (which falls in the ultraviolet (UV) band of the
EM spectrum), the photon’s energy is sufficient to free
an electron and remove it completely, thereby ionizing
the affected atom or molecule. Consequently, the energy
carried by such EM waves is called ionizing radiation,
in contrast with non-ionizing radiation (Fig. TF17-2),
where photons may be able to cause an electron to move
to a higher energy level but not eject it from its host atom
or molecule.
Assessing health risks associated with exposure
to EMFs is complicated by the number of variables
involved, including: (1) the frequency f , (2) the intensities
of the electric and magnetic fields, (3) the exposure
duration—whether continuous or discontinuous, pulsed
or uniform—and (4) the specific part of the body that is
getting exposed. We know that intense laser illumination
can cause corneal burn, high-level X-rays can damage
living tissue and cause cancer, and in fact, any form of
EM energy can be dangerous if the exposure level and/or
duration exceed certain safety limits. Governmental and
professional safety boards are tasked with establishing
maximum permissible exposure (MPE) levels that protect
human beings against adverse health effects associated
with EMFs. In the U.S., the relevant standards are IEEE
Std C95.6 (dated 2002), which addresses EM fields in
the 1 Hz to 3 kHz range, and IEEE Std 95.1 (dated 2005),
which deals with the frequency range from 3 kHz to 300
GHz. On the European side of the Atlantic, responsibility
for establishing MPE levels resides with the Scientific
Committee on Emerging and Newly Identified Health
Risks (SCENIHR) of the European Commission.
◮ At frequencies below 100 kHz, the goal is to
minimize adverse effects of exposure to electric
fields that can cause electrostimulation of nerve
and muscle cells. Above 5 MHz, the main concern
is excessive tissue heating, and in the transition
region of 100 kHz to 5 MHz, safety standards are
designed to protect against both electrostimulation
and excessive heating. ◭
406
CHAPTER 9 RADIATION AND ANTENNAS
Non-ionizing
Ionizing
Extremely low
frequency
Radio
Microwave
Infrared
Visible light
Ultraviolet
X-rays
Gamma rays
Frequency
0.1 Hz
1 MHz
Induces low
currents
1 THz
Induces high
currents
Excites
electrons
1 EHz
Damages DNA
106
Electric Field E (V/m)
10
E-field
Other tissue MPE
105
1
H-field MPE
10–1
10–2
10–3
0.1
E-field
104
103
Brain MPE
1
10
100
1000
Magnetic Field H (A/m)
Figure TF17-2 Different types of electromagnetic radiation.
102
3000
Frequency (Hz)
Figure TF17-3 Maximum permissible exposure (MPE) levels for E and H over the frequency range from 0.1 Hz to 3 kHz.
Frequency Range 0 ≤ f ≤ 3 kHz: The plots in Fig. TF17-3
display the values of MPE for electric and magnetic fields
over the frequency range below 3 kHz. According to
IEEE Std C95.6, it is sufficient to demonstrate compliance with the MPE levels for either the electric field E
or the magnetic field H. According to the plot for H,
exposure at 60 Hz should not exceed 720 A/m. The
magnetic field due to power lines is typically in the range
of 2 to 6 A/m underneath the lines, which is at least two
orders of magnitude smaller than the established safe
level for H.
Frequency Range 3 kHz ≤ f ≤ 300 GHz: At frequencies below 500 MHz, MPE is specified in terms of the
electric and magnetic field strengths of the EM energy
(Fig. TF17-4). From 100 MHz to 300 GHz (and beyond),
MPE is specified in terms of the product of E and H,
namely the power density S. Cell phones operate in the
1 to 2 GHz band; the specified MPE is 1 W/m2 (or
equivalently 0.1 mW/cm2 ).
9-6 FRIIS TRANSMISSION FORMULA
407
10000
10000
1000
614 V/m
Electric field E
163 A/m
100 W/m2 100
100
27.5 V/m
10
10
Power density
2 W/m2
1
Magnetic field H
1
0.0729 A/m
0.1
0.01
10–2
Power Density S (W/m2)
Magnetic Field H (A/m)
Electric Field E (V/m)
1000
0.1
0.01
10–1
1
10
102
103
104
105
Frequency (MHz)
Figure TF17-4 MPE levels for the frequency range from 10 kHz to 300 GHz.
Bottom Line
We are constantly bombarded by EM energy—from
solar illumination to blackbody radiation emitted
by all matter. Our bodies absorb, reflect, and emit
EM energy all the time. Living organisms, including
humans, require exposure to EM radiation to
survive, but excessive exposure can cause adverse
effects. The term excessive exposure connotes
a complicated set of relationships among such
variables as field strength, exposure duration and mode
(continuous, pulsed, etc.), body part, etc. The emission
standards established by the Federal Communications
Commission in the U.S. and similar governmental bodies in other countries are based on a combination of
epidemiological studies, experimental observations, and
theoretical understanding of how EM energy interacts
with biological material. Generally speaking, the maximum permissible exposure levels specified by these
standards are typically two orders of magnitude lower
than the levels known to cause adverse effects, but in
view of the multiplicity of variables involved, there is no
guarantee that adhering to the standards will avoid health
risks absolutely. The bottom line is: Use common sense!
408
CHAPTER 9 RADIATION AND ANTENNAS
Exercise 9-10: The effective area of a parabolic dish
antenna is approximately equal to its physical aperture.
If its directivity is 30 dB at 10 GHz, what is its effective
area? If the frequency is increased to 30 GHz, what will
be its new directivity?
Answer: Ae = 0.07 m2 , D = 39.44 dB. (See
EM
.)
9-7 Radiation by Large-Aperture
Antennas
For wire antennas, the sources of radiation are the infinitesimal
current elements comprising the current distribution along the
wire and the total radiated field at a given point in space is
equal to the sum, or integral, of the fields radiated by all
the elements. A parallel scenario applies to aperture antennas,
except that now the source of radiation is the electric-field
distribution across the aperture. Consider the horn antenna
shown in Fig. 9-20. It is connected to a source through a
coaxial transmission line with the outer conductor of the
line connected to the metal body of the horn and the inner
conductor made to protrude through a small hole partially into
the throat end of the horn. The protruding conductor acts as
a monopole antenna, generating waves that radiate outwardly
toward the horn’s aperture. The electric field of the wave
arriving at the aperture, which may vary as a function of xa
and ya over the horn’s aperture, is called the electric-field aperture distribution or illumination, Ea (xa , ya ). Inside the horn,
wave propagation is guided by the horn’s geometry, but as the
wave transitions from a guided wave into an unbounded wave,
every point of its wavefront serves as a source of spherical
secondary wavelets. The aperture may then be represented
as a distribution of isotropic radiators. At a distant point Q,
the combination of all the waves arriving from all of these
radiators constitutes the total wave that would be observed by
xa
Ea(xa, ya)
R
θ
ya
xa
Ea(xa, ya)
lx
ya
Collimating
lens
ly
(a) Opening in an opaque screen
d
Q
z
Observation
sphere
Figure 9-20 A horn antenna with aperture field distribution
Ea (xa , ya ).
a receiver placed at that point.
The radiation process described for the horn antenna is
equally applicable to any aperture upon which an electromagnetic wave is incident. For example, if a light source is used
to illuminate an opening in an opaque screen through a collimating lens, as shown in Fig. 9-21(a), the opening becomes a
source of secondary spherical wavelets, much like the aperture
of the horn antenna. In the case of the parabolic reflector shown
in Fig. 9-21(b), it can be described in terms of an imaginary
aperture representing the electric-field distribution across a
Imaginary aperture
(b) Parabolic reflector antenna
Figure 9-21 Radiation by apertures: (a) an opening in an
opaque screen illuminated by a light source through a collimating lens and (b) a parabolic dish reflector illuminated by a small
horn antenna.
9-7 RADIATION BY LARGE-APERTURE ANTENNAS
409
practical to construct and use antennas (in this frequency
range) with aperture dimensions that are many wavelengths in
size.
The xa –ya plane in Fig. 9-22 (denoted plane A) contains
an aperture with an electric field distribution Ea (xa , ya ). For
the sake of convenience, the opening has been chosen to
be rectangular in shape, with dimensions lx along xa and ly
along ya , even though the formulation we are about to discuss
is general enough to accommodate any two-dimensional aperture distribution, including those associated with circular and
elliptical apertures. At a distance z from the aperture plane A
in Fig. 9-22, we have an observation plane O with axes (x, y).
The two planes have parallel axes and are separated by a
distance z. Moreover, z is sufficiently large that any point Q in
the observation plane is in the far-field region of the aperture.
To satisfy the far-field condition, it is necessary that
plane in front of the reflector.
Two types of mathematical formulations are available for
computing the electromagnetic fields of waves radiated by
apertures. The first is a scalar formulation based on Kirchhoff’s work, and the second is a vector formulation based on
Maxwell’s equations. In this section, we limit our presentation
to the scalar diffraction technique in part because of its inherent simplicity and also because it is applicable across a wide
range of practical applications.
◮ The key requirement for the validity of the scalar
formulation is that the antenna aperture be at least several wavelengths long along each of its principal dimensions. ◭
A distinctive feature of such an antenna is its high directivity
and correspondingly narrow beam, which makes it attractive
for radar and free-space microwave communication systems.
The frequency range commonly used for such applications is
the 1 to 30 GHz microwave band. Because the corresponding
wavelength range is 30 to 1 cm, respectively, it is quite
R ≥ 2d 2 /λ ,
(far-field range)
(9.73)
where d is the longest linear dimension of the radiating
aperture.
xa
Aperture
illumination
ly
lx
dxa
dya
s
R
z
x
Q
θ
φ
ya
y
Aperture plane A
Observation plane O
Figure 9-22 Radiation by an aperture in the xa –ya plane at z = 0.
410
CHAPTER 9 RADIATION AND ANTENNAS
The position of observation point Q is specified by the
range R between the center of the aperture and point Q and
by the angles θ and φ (Fig. 9-22), which jointly define the
direction of the observation point relative to the coordinate
system of the aperture. In our treatment of the dipole antenna,
we oriented the dipole along the z axis, and we called θ the
zenith angle. In the present context, the z axis is orthogonal
to the plane containing the antenna aperture. Also, θ usually is
called the elevation angle. The electric field phasor of the wave
e θ , φ ). Kirchhoff’s scalar
incident upon point Q is denoted E(R,
diffraction theory provides the following relationship between
e θ , φ ) and the aperture illumination
the radiated field E(R,
e
Ea (xa , ya ):
j e− jkR e
e
E(R, θ , φ ) =
h(θ , φ ),
(9.74)
λ
R
where
e
h(θ , φ ) =
ZZ ∞
−∞
Eea (xa , ya )
· exp[ jk sin θ (xa cos φ + ya sin φ )] dxa dya . (9.75)
e θ , φ ).
We shall refer to e
h(θ , φ ) as the form factor of E(R,
Its integral is written with infinite limits with the understanding that Eea (xa , ya ) is identically zero outside the aperture. The spherical propagation factor (e− jkR/R) accounts for
wave propagation between the center of the aperture and the
observation point, and e
h(θ , φ ) represents an integration of the
exciting field Eea (xa , ya ) over the extent of the aperture, taking
into account [through the exponential function in Eq. (9.75)]
the approximate deviation in distance between R and s, where
s is the distance to any point (xa , ya ) in the aperture plane (see
Fig. 9-22).
◮ In Kirchhoff’s scalar formulation, the polarization
e θ , φ ) is the same as
direction of the radiated field E(R,
that of the aperture field Eea (xa , ya ). ◭
Also, the power density of the radiated wave is given by
S(R, θ , φ ) =
e θ , φ )|2
|E(R,
|e
h(θ , φ )|2
.
=
2η0
2η0 λ 2 R2
(9.76)
9-8 Rectangular Aperture with Uniform
Aperture Distribution
To illustrate the scalar diffraction technique, consider a rectangular aperture of height lx and width ly , both at least a few
wavelengths long. The aperture is excited by a uniform field
distribution (i.e., constant value) given by

 E0 for − lx /2 ≤ xa ≤ lx /2
and − ly /2 ≤ ya ≤ ly /2,
Eea (xa , ya ) =
(9.77)
 0
otherwise.
To keep the mathematics simple, let us confine our examination to the radiation pattern at a fixed range R in the
x–z plane, which corresponds to φ = 0. In this case, Eq. (9.75)
simplifies to
e
h(θ ) =
Z lx /2
Z ly /2
ya =−ly /2 xa =−lx /2
E0 exp[ jkxa sin θ ] dxa dya . (9.78)
In preparation for performing the integration in Eq. (9.78), we
introduce the intermediate variable u defined as
2π sin θ
u = k sin θ =
.
(9.79)
λ
Hence,
e
h(θ ) = E0
Z lx /2
−lx /2
e juxa dxa ·
Z ly /2
−ly /2
dya
#
e julx /2 − e− julx /2
· ly
= E0
ju
"
#
2E0 ly e julx /2 − e− julx /2
2E0 ly
=
=
sin(ulx /2).
u
2j
u
(9.80)
"
Upon replacing u with its defining expression, we have
2E0 ly
e
sin(π lx sin θ /λ )
h(θ ) = 2π
sin θ
λ
= E0 lx ly
sin(π lx sin θ /λ )
π lx sin θ /λ
= E0 Ap sinc(π lx sin θ /λ ),
(9.81)
where Ap = lx ly is the physical area of the aperture. Also, we
used the standard definition of the sinc function, which for any
argument t is defined as
sint
.
(9.82)
t
Using Eq. (9.76), we obtain the following expression for the
power density at the observation point:
sinct =
S(R, θ ) = S0 sinc2 (π lx sin θ /λ ) (x–z plane),
where S0 = E02 A2p /(2η0 λ 2 R2 ).
(9.83)
9-8 RECTANGULAR APERTURE WITH UNIFORM APERTURE DISTRIBUTION
−3 dB on a decibel scale), as shown in Fig. 9-23. Since the
pattern is symmetrical with respect to θ = 0, θ1 = −θ2 , and
βxz = 2θ2 . The angle θ2 can be obtained from a solution of
F(γ)
0 dB
z
Q = (R, θ)
R y
a
θ
ly
xa
F(θ2 ) = sinc2 (π lx sin θ /λ ) = 0.5.
–3
From tabulated values of the sinc function, it is found that
Eq. (9.85) yields the result
–10
π lx
sin θ2 = 1.39,
λ
(9.86)
λ
.
lx
(9.87)
βxz
–13.2 dB
or
–15
sin θ2 = 0.44
Because λ /lx ≪ 1 (a fundamental condition of scalar diffraction theory is that the aperture dimensions be much larger than
the wavelength λ ), θ2 is a small angle, in which case we can
use the approximation sin θ2 ≈ θ2 . Hence,
–20
–25
–2
(9.85)
–5
lx
–3
411
–30
–1
1
γ = (lx/λ) sin θ
βxz = 2θ2 ≈ 2 sin θ2 = 0.88
2
3
λ
lx
(rad).
(9.88a)
A similar solution for the y–z plane (φ = π /2) gives
Figure 9-23 Normalized radiation pattern of a uniformly
illuminated rectangular aperture in the x–z plane (φ = 0).
βyz = 0.88
◮ The sinc function is maximum when its argument is
zero; sinc(0) = 1. ◭
This occurs when θ = 0. Hence, at a fixed range R,
Smax = S(θ = 0) = S0 . The normalized radiation intensity is
then given by
S(R, θ )
= sinc2 (π lx sin θ /λ )
F(θ ) =
Smax
= sinc2 (πγ )
(x–z plane).
(9.84)
Figure 9-23 shows F(θ ) plotted (on a decibel scale) as a
function of the intermediate variable γ = (lx /λ ) sin θ . The
pattern exhibits nulls at nonzero integer values of γ .
9-8.1 Beamwidth
The normalized radiation intensity F(θ ) is symmetrical in the
x–z plane, and its maximum is along the boresight direction
(θ = 0, in this case). Its half-power beamwidth βxz = θ2 − θ1 ,
where θ1 and θ2 are the values of θ at which F(θ , 0) = 0.5 (or
λ
ly
(rad).
(9.88b)
◮ The uniform aperture distribution (Eea = E0 across
the aperture) gives a far-field pattern with the narrowest
possible beamwidth. ◭
The first sidelobe level is 13.2 dB below the peak value
(see Fig. 9-23), which is equivalent to 4.8% of the peak
value. If the intended application calls for a pattern with a
lower sidelobe level (to avoid interference with signals from
sources along directions outside the main beam of the antenna
pattern), this can be accomplished by using a tapered aperture
distribution—one that is a maximum at the center of the
aperture and decreases toward the edges.
◮ A tapered distribution provides a pattern with lower
side lobes, but the main lobe becomes wider. ◭
The steeper the taper, the lower are the side lobes and the wider
is the main lobe. In general, the beamwidth in a given plane,
412
CHAPTER 9 RADIATION AND ANTENNAS
single major lobe whose boresight is along the z direction:
Sidelobes
β≈ λ
d
d
Boresight
4π
.
βxz βyz
βxz ≈ λ
lx
βyz ≈ λ
ly
◮ For aperture antennas, their effective apertures are
approximately equal to their physical apertures; that is,
Ae ≈ Ap . ◭
lx
Exercise 9-11: Verify that Eq. (9.86) is a solution of
Eq. (9.85) by calculating sinc2 t for t = 1.39.
(b) Fan beam
Figure 9-24 Radiation patterns of (a) a circular reflector and
(b) a cylindrical reflector (side lobes not shown).
say the x–z plane, is given by
Exercise 9-12: With its boresight direction along z,
a square aperture was observed to have half-power
beamwidths of 3◦ in both the x–z and y–z planes. Determine its directivity in decibels.
Answer: D = 4,583.66 = 36.61 dB. (See
βxz = kx
(9.90)
If we use the approximate relations βxz ≈ λ /lx and βyz ≈ λ /ly ,
we obtain
4π lx ly
4 π Ap
D≈
=
.
(9.91)
λ2
λ2
For any antenna, its directivity is related to its effective area Ae
by Eq. (9.64):
4 π Ae
.
(9.92)
D=
λ2
(a) Pencil beam
ly
D≈
λ
,
lx
(9.89)
where kx is a constant related to the steepness of the taper. For
a uniform distribution with no taper, kx = 0.88, and for a highly
tapered distribution, kx ≈ 2. In the typical case, kx ≈ 1.
To illustrate the relationship between the antenna dimensions and the corresponding beam shape, we show in Fig. 9-24
the radiation patterns of a circular reflector and a cylindrical
reflector. The circular reflector has a circularly symmetric
pattern, whereas the pattern of the cylindrical reflector has a
narrow beam in the azimuth plane corresponding to its long dimension and a wide beam in the elevation plane corresponding
to its narrow dimension. For a circularly symmetric antenna
pattern, the beamwidth β is related to the diameter d by the
approximate relation β ≈ λ /d.
9-8.2 Directivity and Effective Area
In Section 9-2.3, we derived an approximate expression
[Eq. (9.26)] for the antenna directivity D in terms of the halfpower beamwidths βxz and βyz for antennas characterized by a
EM
.)
Exercise 9-13: What condition must be satisfied in order
to use scalar diffraction to compute the field radiated
by an aperture antenna? Can we use it to compute the
directional pattern of the eye’s pupil (d ≈ 0.2 cm) in the
visible part of the spectrum (λ = 0.35 to 0.7 µ m)? What
would the beamwidth of the eye’s directional pattern be at
λ = 0.5 µ m?
Answer: β ≈ λ /d = 2.5 × 10−4 rad = 0.86′ (arc minute,
with 60′ = 1◦ ). (See EM .)
9-9 Antenna Arrays
AM broadcast services operate in the 535 to 1605 kHz band.
The antennas they use are vertical dipoles mounted along
tall towers. The antennas range in height from λ /6 to 5λ /8,
depending on the operating characteristics desired and other
considerations. Their physical heights vary from 46 m (150 ft)
to 274 m (900 ft); the wavelength at 1 MHz, approximately
in the middle of the AM band, is 300 m. Because the field
radiated by a single dipole is uniform in the horizontal plane
9-9 ANTENNA ARRAYS
413
Module 9.4 Large Parabolic Reflector For any specified reflector diameter d (where d ≥ 2λ ) and illumination taper
factor α , this module displays the pattern of the radiated field and computes the associated beamwidth and directivity.
(as discussed in Sections 9-1 and 9-3), it is not possible to direct the horizontal pattern along specific directions of interest,
unless two or more antenna towers are used simultaneously.
Directions of interest may include cities serviced by the AM
station, and directions to avoid may include areas serviced
by another station operating at the same frequency (thereby
avoiding interference effects). When two or more antennas are
used together, the combination is called an antenna array.
The AM broadcast antenna array is only one example of the
many antenna arrays used in communication systems and radar
applications. Antenna arrays provide the antenna designer the
flexibility to obtain high directivity, narrow beams, low side
lobes, steerable beams, and shaped antenna patterns starting
from very simple antenna elements. Figure 9-25 shows a very
large radar system consisting of a transmitter array composed
of 5,184 individual dipole antenna elements and a receiver
array composed of 4,660 elements. The radar system, part
of the Space Surveillance Network operated by the U.S. Air
Force, operates at 442 MHz and transmits a combined peak
power of 30 MW!
Although an array need not consist of similar radiating
elements, most arrays actually use identical elements, such
as dipoles, slots, horn antennas, or parabolic dishes. The
antenna elements comprising an array may be arranged in
various configurations, but the most common are the linear one-dimensional configuration—wherein the elements are
arranged along a straight line—and the two-dimensional lattice
configuration in which the elements sit on a planar grid. The
desired shape of the far-field radiation pattern of the array can
be synthesized by controlling the relative amplitudes of the
array elements’ excitations.
◮ A
Download